Open source AI is the path forward

https://about.fb.com/news/2024/07/open-source-ai-is-the-path-forward/

by atgctg • 2 years ago

2360 872 2 years ago

about.fb.com

Sure but under what license? Because slapping “open source” on the model doesn’t make it open source if it’s not actually license that way. The 3.1 license still contains their non-commercial clause (over 700m users) and requires derivatives, whether fine tunings or trained on generated data, to use the llama name.

redleader55• 2 years ago

"Use it for whatever you want(conditions apply), but not if you are Google, Amazon, etc. If you become big enough talk to us." That's how I read the license, but obviously I might be missing some nuance.

mesebrec8• 2 years ago

You also can't use it for training or improving other models.

You also can't use it if you're the government of India.

Neither can sex workers use it. (Do you know if your customers are sex workers?)

There are also very vague restrictions for things like discrimination, racism etc.

frabcus10• 2 years ago

Also it isn't source code, it is a binary. You need at least the data curation code and preferably the data itself for it to be actually source code in the practical sense that anyone can remake the build.

Llama could change the license on later versions to kill your business and you have no options as you don't know how they trained it or have the budget to.

It's not much more free than binary software.

aliljet• 2 years ago

And this is happening RIGHT as a new potential leader is emerging in Llama 3.1. I'm really curious about how this is going to match up on the leaderboards...

kart23• 2 years ago

This is how we’ve managed security on our social networks – our more robust AI systems identify and stop threats from less sophisticated actors who often use smaller scale AI systems.

Ok, first of all, has this really worked? AI moderators still can't capture the mass of obvious spam/bots on all their platforms, threads included. Second, AI detection doesn't work, and with how much better the systems are getting, it's probably never going to, unless you keep the best models for yourself, and it's is clear from the rest of the note that its not zuck's intention to do so.

As long as everyone has access to similar generations of models – which open source promotes – then governments and institutions with more compute resources will be able to check bad actors with less compute.

This just doesn't make sense. How are you going to prevent AI spam, AI deepfakes from causing harm with more compute? What are you gonna do with more compute about nonconsensual deepfakes? People are already using AI to bypass identity verification on your social media networks, and pump out loads of spam.

OpenComment• 2 years ago

Interesting quotes. Less sophisticated actors just means humans who already write in 2020 what the NYT wrote in early 2022 to prepare for Biden's State Of The Union 180° policy reversals (manufacturing consent).

FB was notorious for censorship. Anyway, what is with the "actions/actors" terminology? This is straightforward totalitarian language.

simonw2• 2 years ago

"AI detection doesn't work, and with how much better the systems are getting, it's probably never going to, unless you keep the best models for yourself"

I don't think that's true. I don't think even the best privately held models will be able to detect AI text reliably enough for that to be worthwhile.

zmmmmm5• 2 years ago

I found this dubious as well, especially how it is portrayed as a simple game of compute power. For a start, there is an enormous asymmetry which is why we have a spam problem in the first place. For example a single bot can send out millions of emails at almost no cost and we have to expend a lot more "energy" to classify each one and decide if it's spam or not. So you don't just need more compute power you need drastically more compute power, and as AI models improve and get refined, the operation at ten times the scale is probably going to be marginally better, not orders of magnitude better.

I still agree with his general take - bad actors will get these models or make them themselves, you can't stop it. But the logic about compute power is odd.

blackeyeblitzar9• 2 years ago

Only if it is truly open source (open data sets, transparent curation/moderation/censorship of data sets, open training source code, open evaluation suites, and an OSI approved open source license).

Open weights (and open inference code) is NOT open source, but just some weak open washing marketing.

The model that comes closest to being TRULY open is AI2’s OLMo. See their blog post on their approach:

https://blog.allenai.org/hello-olmo-a-truly-open-llm-43f7e73...

I think the only thing they’re not open about is how they’ve curated/censored their “Dolma” training data set, as I don’t think they explicitly share each decision made or the original uncensored dataset:

https://blog.allenai.org/dolma-3-trillion-tokens-open-llm-co...

By the way, OSI is working on defining open source for AI. They post weekly updates to their blog. Example:

https://opensource.org/blog/open-source-ai-definition-weekly...

JumpCrisscross21• 2 years ago

Only if it is truly open source (open data sets, transparent curation/moderation/censorship of data sets, open training source code, open evaluation suites, and an OSI approved open source license)

You’re missing a then to your if. What happens if it’s “truly” open per your definition versus not?

blackeyeblitzar9• 2 years ago

I think you are asking what the benefits are? The main benefit is that we can trust what these systems are doing better. Or we can self host them. If we just take the weights, then it is unclear how these systems might be lying to us or manipulating us.

Another benefit is that we can learn from how the training and other steps actually work. We can change them to suit our needs (although costs are impractical today). Etc. It’s all the usual open source benefits.

haolez• 2 years ago

There is also the risk of companies like Meta introducing ads in the training itself, instead of inference time.

itissid2• 2 years ago

Yeah, though I do wonder for a big model like 405B if the original training recipe, really matters for where models are heading, practically speaking which is smaller and more specific?

I imagine its main use would be to train other models by distilling them down with LoRA/Quantization etc(assuming we have a tokenizer). Or use them to generate training data for smaller models directly.

But, I do think there is always a way to share without disclosing too many specifics, like this[1] lecture from this year's spring course at Stanford. You can always say, for example:

- The most common technique for filtering was using voting LLMs (without disclosing said llms or quantity of data).

- We built on top of a filtering technique for removing poor code using ____ by ____ authors (without disclosing or handwaving how you exactly filtered, but saying that you had to filter).

- We mixed certain proportion of this data with that data to make it better (without saying what proportion)

[1] https://www.youtube.com/watch?v=jm2hyJLFfN8&list=PLoROMvodv4...

JumpCrisscross21• 2 years ago

“The Heavy Press Program was a Cold War-era program of the United States Air Force to build the largest forging presses and extrusion presses in the world.” This ”program began in 1944 and concluded in 1957 after construction of four forging presses and six extruders, at an overall cost of $279 million. Six of them are still in operation today, manufacturing structural parts for military and commercial aircraft” [1].

$279mm in 1957 dollars is about $3.2bn today [2]. A public cluster of GPUs provided for free to American universities, companies and non-profits might not be a bad idea.

[1] https://en.m.wikipedia.org/wiki/Heavy_Press_Program

[2] https://data.bls.gov/cgi-bin/cpicalc.pl?cost1=279&year1=1957...

CardenB2• 2 years ago

Doubtful that GPUs purchased today would be in use for a similar time scale. Govt investment would also drive the cost of GPUs up a great deal.

Not sure why a publicly accessible GPU cluster would be a better solution than the current system of research grants.

JumpCrisscross21• 2 years ago

Doubtful that GPUs purchased today would be in use for a similar time scale

Totally agree. That doesn't mean it can't generate massive ROI.

Govt investment would also drive the cost of GPUs up a great deal

Difficult to say this ex ante. On its own, yes. But it would displace some demand. And it could help boost chip production in the long run.

Not sure why a publicly accessible GPU cluster would be a better solution than the current system of research grants

Those receiving the grants have to pay a private owner of the GPUs. That gatekeeping might be both problematic, if there is a conflict of interests, and inefficient. (Consider why the government runs its own supercomputers versus contracting everything to Oracle and IBM.)

ygjb3• 2 years ago

Of course they won't. The investment in the Heavy Press Program was the initial build, and just citing one example, the Alcoa 50,000 ton forging press was built in 1955, operated until 2008, and needed ~$100M to get it operational again in 2012.

The investment was made to build the press, which created significant jobs and capital investment. The press, and others like it, were subsequently operated by and then sold to a private operator, which in turn enabled the massive expansion of both military manufacturing, and commercial aviation and other manufacturing.

The Heavy Press Program was a strategic investment that paid dividends by both advancing the state of the art in manufacturing at the time it was built, and improving manufacturing capacity.

A GPU cluster might not be the correct investment, but a strategic investment in increasing, for example, the availability of training data, or interoperability of tools, or ease of use for building, training, and distributing models would probably pay big dividends.

jvanderbot• 2 years ago

A much better investment would be to (somehow) revolutionize production of chips for AI so that it's all cheaper, more reliable, and faster to stand up new generations of software and hardware codesign. This is probably much closer to the program mentioned in the top level comment: It wasn't to produce one type of thing, but to allow better production of any large thing from lighter alloys.

photonthug2• 2 years ago

Not sure why a publicly accessible GPU cluster would be a better solution than the current system of research grants.

You mean a better solution than different teams paying AWS over and over, potentially spending 10x on rent rather than using all that cash as a down payment on actually owning hardware? I can't really speak for the total costs of depreciation/hardware maintenance but renting forever isn't usually a great alternative to buying.

light_hue_1• 2 years ago

The problem is that any public cluster would be outdated in 2 years. At the same time, GPUs are massively overpriced. Nvidia's profit margins on the H100 are crazy.

Until we get cheaper cards that stand the test of time, building a public cluster is just a waste of money. There are far better ways to spend $1b in research dollars.

JumpCrisscross21• 2 years ago

any public cluster would be outdated in 2 years

The private companies buying hundreds of billions of dollars of GPUs aren't writing them off in 2 years. They won't be cutting edge for long. But that's not the point--they'll still be available.

Nvidia's profit margins on the H100 are crazy

I don't see how the current practice of giving a researcher a grant so they can rent time on a Google cluster that runs H100s is more efficient. It's just a question of capex or opex. As a state, the U.S. has a structual advantage in the former.

far better ways to spend $1b in research dollars

One assumes the U.S. government wouldn't be paying list price. In any case, the purpose isn't purely research ROI. Like the heavy presses, it's in making a prohibitively-expensive capital asset generally available.

ninininino• 2 years ago

What about dollar cost averaging your purchases of GPUs? So that you're always buying a bit of the newest stuff every year rather than just a single fixed investment in hardware that will become outdated? Say 100 million a year every year for 20 years instead of 2 billion in a single year?

fweimer• 2 years ago

Don't these public clusters exist today, and have been around for decades at this point, with varying architectures? In the sense that you submit a proposal, it gets approved, and then you get access for your research?

JumpCrisscross21• 2 years ago

Not--to my knowledge--for the GPUs necessary to train cutting-edge LLMs.

NewJazz• 2 years ago

This is the most recent iteration of a national platform. They have tons of GPUs (and CPUs, and flash storage) hooked up as a Kubernetes cluster, available for teaching and research.

https://nationalresearchplatform.org/

epaulson• 2 years ago

The National Science Foundation has been doing this for decades, starting with the supercomputing centers in the 80s. Long before anyone talked about cloud credits, NSF has had a bunch of different programs to allocate time on supercomputers to researchers at no cost, these days mostly run out of the Office of Advanced Cyberinfrastruture. (The office name is from the early 00s) - https://new.nsf.gov/cise/oac

(To connect universities to the different supercomputing centers, the NSF funded the NSFnet network in the 80s, which was basically the backbone of the Internet in the 80s and early 90s. The supercomputing funding has really, really paid off for the USA)

JumpCrisscross21• 2 years ago

NSF has had a bunch of different programs to allocate time on supercomputers to researchers at no cost, these days mostly run out of the Office of Advanced Cyberinfrastruture

This would be the logical place to put such a programme.

jszymborski3• 2 years ago

As you've rightly pointed out, we have the mechanism, now let's fund it properly!

I'm in Canada, and our science funding has likewise fallen year after year as a proportion of our GDP. I'm still benefiting from A100 clusters funded by tax payer dollars, but think of the advantage we'd have over industry if we didn't have to fight over resources.

cmdrk• 2 years ago

Yeah, the specific AI/ML-focused program is NAIRR.

https://nairrpilot.org/

Terrible name unless they low-key plan to make AI researchers' hair fall out.

dastbe• 2 years ago

the US already pays for 2+ aws region for cia/dod. why not pay for a region that is only available to researchers?

blackeyeblitzar9• 2 years ago

What about distributed training on volunteer hardware? Is that feasible?

oersted2• 2 years ago

It is an exciting concept, there's a huge wealth of gaming hardware deployed that is inactive at most hours of the day. And I'm sure people are willing to pay well above the electricity cost for it.

Unfortunately, the dominant LLM architecture makes it relatively infeasible right now.

- Gaming hardware has too limited VRAM for training any kind of near-state-of-the-art model. Nvidia is being annoyingly smart about this to sell enterprise GPUs at exorbitant markups.

- Right now communication between machines seems to be the bottleneck, and this is way worse with limited VRAM. Even with data-centre-grade interconnect (mostly Infiniband, which is also Nvidia, smart-asses), any failed links tend to cause big delays in training.

Nevertheless, it is a good direction to push towards, and the government could indeed help, but it will take time. We need both a more healthy competitive landscape in hardware, and research towards model architectures that are easy to train in a distributed manner (this was also the key to the success of Transformers, but we need to go further).

codemusings• 2 years ago

Ever heard of SETI@home?

https://setiathome.berkeley.edu

ks2048• 2 years ago

How about using some of that money to develop CUDA alternatives so everyone is not paying the Nvidia tax?

lukan• 2 years ago

It would be probably cheaper to negate some IP. There are quite some projects and initiatives to make CUDA code run on AMD for example, but as far as I know, they all stopped at some point, probably because of fear of being sued into oblivion.

whimsicalism13• 2 years ago

It seems like rocm is already fully ready for transformer inference, so you are just referring to training?

belter• 2 years ago

Please start with the Windows Tax first for Linux users buying hardware...and the Apple Tax for Android users...

zitterbewegung• 2 years ago

Either you port Tensorflow (Apple)[1] or PyTorch to your platform or you allow CUDA to run on your hardware (AMD) [2]. Companies are incentives to not have NVIDIA having a monopoly but the thing is that CUDA is a huge moat due to compatibility of all frameworks and everyone knows it. Also, all of the cloud or on premises providers use NVIDIA regardless.

[1] https://developer.apple.com/metal/tensorflow-plugin/ [2] https://www.xda-developers.com/nvidia-cuda-amd-zluda/

erickj• 2 years ago

That's the kind of work that can come out of academia and open source communities when societies provide the resources required.

latchkey3• 2 years ago

It is being done already...

https://docs.scale-lang.com/

dogcomplex3• 2 years ago

Or just develop the next wave of chips designed for specifically transformer-based architectures (and ternary computing), and bypass the needs for GPUs and CUDA altogether

prpl• 2 years ago

Great idea, too bad the DOE and NSF were there first.

kjkjadksj2• 2 years ago

The size of the cluster would have to be massive or else your job will be on the queue for a year. And even then what are you going to do downsize the resources requested so you can get in earlier? After a certain point it starts to make more sense to just buy your own xeons and run your own cluster.

Aperocky• 2 years ago

Imagine if they made a data center with 1957 electronics that cost $279 million.

They probably won't be using it now because the phone in your pocket is likely more powerful. Moore law did end but data center stuff are still evolving order of magnitudes faster than forging presses.

goda90• 2 years ago

I'd like to see big programs to increase the amount of cheap, clean energy we have. AI compute would be one of many beneficiaries of super cheap energy, especially since you wouldn't need to chase newer, more efficient hardware just to keep costs down.

Melatonic• 2 years ago

Yeah this would be the real equivalent of the program people are talking about above. That an investing in core networking infrastructure (like cables) instead of just giving huge handouts to certain corporations that then pocket the money.....

BigParm• 2 years ago

So we'll have the government bypass markets and force the working class to buy toys for the owning class?

If anything, allocate compute to citizens.

_fat_santa• 2 years ago

If anything, allocate compute to citizens.

If something like this were to become a reality, I could see something like "CitizenCloud" where once you prove that you are a US Citizen (or green card holder or some other requirement), you can then be allocated a number of credits every month for running workloads on the "CitizenCloud". Everyone would get a baseline amount, from there if you can prove you are a researcher or own a business related to AI then you can get more credits.

aiauthoritydev• 2 years ago

Overall government doing anything is a bad idea. There are cases however where government is the only entity that can do certain things. These are things that involve military, law enforcement etc. Outside of this we should rely on private industry and for-profit industry as much as possible.

pavlov• 2 years ago

The American healthcare industry demonstrates the tremendous benefits of rigidly applying this mindset.

Why couldn’t law enforcement be private too? You call 911, several private security squads rush to solve your immediate crime issue, and the ones who manage to shoot the suspect send you a $20k bill. Seems efficient. If you don’t like the size of the bill, you can always get private crime insurance.

chris_wot• 2 years ago

That’s not correct. The American health care system is an extreme example of where private organisations fail overall society.

fragmede• 2 years ago

Overall government doing anything is a bad idea.

that is bereft of detail enough to just be wrong. There are things that government is good for and things that government is bad for, but "anything" is just too broad, and reveals an anti-government bias which just isn't well thought out.

goatlover2• 2 years ago

Why are governments a bad idea? Seems the human race has opted for governments doing things since the dawn of civilization. Building roads, providing defense, enforcing rights, provide social safety nets, funding costly scientific endeavors.

com2kid• 2 years ago

Ugh.

Government distorting undeveloped markets that have a lot of room for competition to increase efficiencies is a bad thing.

Government agencies running programs that should not be profitable, or where the only profit to be left comes at the expense of society as a whole, is a good thing.

Lots of basic medicine is the go to example here, treating cancer isn't going to be "profitable" and attempting to make it such just leads to dead people.

On the flip side, one can argue that dentistry has seen amazing strides in affordability and technological progress through the free market. From dental xrays to improvements in dental procedures to make them less painful for the patients.

Eye surgery is another area where competition has lead to good consumer outcomes.

But life of death situations where people can't spend time researching? The only profit there comes through exploiting people.

Angostura• 2 years ago

To summarise: There are some things where government action is the best solution, however by default see if the private sector can sort it first.

varenc• 2 years ago

I just watched this 1950s DoD video on the heavy press program and highly recommend it: https://www.youtube.com/watch?v=iZ50nZU3oG8

newzisforsukas• 2 years ago

https://loc.gov/pictures/search/?q=Photograph:%20oh1540&fi=n...

spullara• 2 years ago

It makes much more sense to invest in a next generation fab for GPUs than to buy GPUs and more closely matches this kind of project.

epolanski• 2 years ago

Does it? You're looking at a gargantuan investment in terms of money that would also require thousands of staff.

That just doesn't seem a good idea.

whartung• 2 years ago

Now, I have no idea.

How much capability would $3.2bn in terms of AI computing power provide, including the operational and power costs of the cluster?

Certainly, you could build a "$3.2bn GPU cluster", but it would be dark.

So, how much learning time would $3.2bn provide? 1 year? 10 years?

Just curious about hand wavy guesses. I have no idea the scope of the these clusters.

rkique• 2 years ago

Very much in this spirit is the NSF-funded National Deep Inference Fabric, which lets researchers run remote experiments on foundation models: https://ndif.us. They just announced a pilot program for Llama405b!

cyanydeez• 2 years ago

Better idea would be to make various open source packages utilities and put maintainers everywhere funded by public good.

AI is a fad, the brick and mortar of the future is open source tools.

fintler• 2 years ago

For the DoE, take a look at:

https://doeleadershipcomputing.org/

carschno• 2 years ago

In the Netherlands, for instance, there is "the national supercomputer" Snellius: https://www.surf.nl/en/services/snellius-the-national-superc... I am not sure about its budget, but my impression as a user is that its resources are never fully used. At least, none of my jobs ever had to queue. I doubt that it can compete with the scale of resources that FAANG companies have available, but then again, I also doubt how research would benefit.

Sure, academia could build LLMs, and there is at least one large-scale project for that: https://gpt-nl.com/ On the other hand, this kind of models still need to demonstrate specific scientific value that goes beyond using a chatbot for generating ideas and summarizing documents.

So I fully agree that the research budget cuts in the past decades have been catastrophic, and probably have contributed to all the disasters the world is currently facing. But I think that funding prestigious super-projects is not the best way to spend funds.

teekert• 2 years ago

Snellius is a nice resource. A powerful Slurm based HTC cluster with different cues for different workloads (cpu/genomics, gpu/deep learning).

To access the resource I had to go through EuroCC [0], which is a network facilitating access to and exploitation of HPC/HTC infra. It is (or can be) a great competing model to US cloud providers.

As a small business I got 8 hrs of consultancy and 10k compute hours for free. I’m still learning the details but my understanding is is that after that the prices are very competitive.

[0] https://www.eurocc-access.eu/

matteocontrini• 2 years ago

Italy built the Leonardo HPC cluster, it's one of the largest in EU and was created by a consortium of universities. After just over a year it's already at full capacity and expansion plans have been anticipated because of this.

B4CKlash• 2 years ago

Eric Schmidt advocated for this exact thing in an Op-ed piece in the latest MIT Technology Review.

[1] https://www.technologyreview.com/2024/05/13/1092322/why-amer...

bayindirh• 2 years ago

A public cluster of GPUs provided for free to American universities, companies and non-profits might not be a bad idea.

USA and Europe is already doing that in a grand scale, in different forms. Both at national and international scale.

I work at an HPC center which provides servers nationally and collaborates on international level.

ldjkfkdsjnv• 2 years ago

On every post like this on hacker news, I post my pro meta speech. My opinion has always been the same:

1. Meta pushed engineering wages higher across the industry.

2. They promote high performing engineers very quickly. There are engineers making 7 figures there with just a few years experience.

3. They have open sourced the most important frameworks: React and Pytorch

This company is a guiding light forcing the hand of other large corporations. Mark Zuckerberg is a hero, and has done a fantastic job

LorenDB• 2 years ago

I never really liked Zuck due to Facebook's massive black hole of user tracking, but you've gotta give it to him: the Llama models are insanely good, and the commitment to making Horizon OS available to other VR manufacturers is commendable, although I'd like to see it truly open sourced.

dang2• 2 years ago

Ok, but please don't post repetitive comments or recite speeches on HN. We're trying to avoid that kind of thing here, because repetition is tedious and bad for curiosity.

https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...

https://news.ycombinator.com/newsguidelines.html

maxdo3• 2 years ago

so that North Korea will create small call centers for cheaper, since they can get these models for free?

HanClinto• 2 years ago

The article argues that the threat of foreign espionage is not solved by closing models.

Some people argue that we must close our models to prevent China from gaining access to them, but my view is that this will not work and will only disadvantage the US and its allies. Our adversaries are great at espionage, stealing models that fit on a thumb drive is relatively easy, and most tech companies are far from operating in a way that would make this more difficult. It seems most likely that a world of only closed models results in a small number of big companies plus our geopolitical adversaries having access to leading models, while startups, universities, and small businesses miss out on opportunities.

tempfile3• 2 years ago

This argument implies that cheap phones are bad since telemarketers can use them.

mrfinn• 2 years ago

You guys really need to get over your bellicose POV of the world. Actually, before it destroys you. Really, is not necessary. Most people in the world just want to leave in peace, and see their children grow happily. For each data center NK would create there will be a thousand of peaceful, kind, and well-intentioned AI projects going on. Or maybe more.

the8thbit20• 2 years ago

"Eventually though, open source Linux gained popularity – initially because it allowed developers to modify its code however they wanted ..."

I find the language around "open source AI" to be confusing. With "open source" there's usually "source" to open, right? As in, there is human legible code that can be read and modified by the user? If so, then how can current ML models be open source? They're very large matrices that are, for the most part, inscrutable to the user. They seem akin to binaries, which, yes, can be modified by the user, but are extremely obscured to the user, and require enormous effort to understand and effectively modify.

"Open source" code is not just code that isn't executed remotely over an API, and it seems like maybe its being conflated with that here?

orthoxerox2• 2 years ago

Open training dataset + open steps sufficient to train exactly the same model.

the8thbit20• 2 years ago

This isn't what Meta releases with their models, though I would like to see more public training data. However, I still don't think that would qualify as "open source". Something isn't open source just because its reproducible out of composable parts. If one, very critical and system defining part is a binary (or similar) without publicly available source code, then I don't think it can be said to be "open source". That would be like saying that Windows 11 is open source because Windows Calculator is open source, and its a component of Windows.

Yizahi2• 2 years ago

They can't release training dataset if it was illegally scrapped all over the web without permission :) (taps head)

bilsbie2• 2 years ago

Can’t you do fine tuning on those binaries? That’s a modification.

the8thbit20• 2 years ago

You can fine tune the models, and you can modify binaries. However, there is no human readable "source" to open in either case. The act of "fine tuning" is essentially brute forcing the system to gradually alter the weights such that loss is reduced against a new training set. This limits what you can actually do with the model vs an actual open source system where you can understand how the system is working and modify specific functionality.

Additionally, models can be (and are) fine tuned via APIs, so if that is the threshold required for a system to be "open source", then that would also make the GPT4 family and other such API only models which allow finetuning open source.

jsheard• 2 years ago

I also think that something like Chromium is a better analogy for corporate open source models than a grassroots project like Linux is. Chromium is technically open source, but Google has absolute control over the direction of it's development and realistically it's far too complex to maintain a fork without Googles resources, just like Meta has complete control over what goes into their open models, and even if they did release all the training data and code (which they don't) us mere plebs could never afford to train a fork from scratch anyway.

skybrian3• 2 years ago

I think you’re right from the perspective of an individual developer. You and I are not about to fork Chromium any time soon. If you presume that forking is impractical then sure, the right to fork isn’t worth much.

But just because a single developer couldn’t do it doesn’t mean it couldn’t be done. It means nobody has organized a large enough effort yet.

For something like a browser, which is critical for security, you need both the organization and the trust. Despite frequent criticism, Mozilla (for example) is still considered pretty trustworthy in a way that an unknown developer can’t be.

candiddevmike• 2 years ago

None of Meta's models are "open source" in the FOSS sense, even the latest Llama 3.1. The license is restrictive. And no one has bothered to release their training data either.

This post is an ad and trying to paint these things as something they aren't.

JumpCrisscross21• 2 years ago

no one has bothered to release their training data

If the FOSS community sets this as the benchmark for open source in respect of AI, they're going to lose control of the term. In most jurisdictions it would be illegal for the likes of Meta to release training data.

blackeyeblitzar9• 2 years ago

AI2 has released training data in their OLMo model: https://blog.allenai.org/hello-olmo-a-truly-open-llm-43f7e73...

causal4• 2 years ago

"Open weights" is a more appropriate term but I'll point out that these weights are also largely inscrutable to the people with the code that trained it. And for licensing reasons, the datasets may not be possible to share.

There is still a lot of modifying you can do with a set of weights, and they make great foundations for new stuff, but yeah we may never see a competitive model that's 100% buildable at home.

Edit: mkolodny points out that the model code is shared (under llama license at least), which is really all you need to run training https://github.com/meta-llama/llama3/blob/main/llama/model.p...

aerzen• 2 years ago

LLAMA is an open-weights model. I like this term, let's use that instead of open source.

stavros4• 2 years ago

"Open weights" means you can use the weights for free (as in beer). "Open source" means you get the training dataset and the methodology. ~Nobody does open source LLMs.

ab5tract4• 2 years ago

If you can’t share the dataset, under what twisted reality are you fine to share the derivative models based on those unsharable datasets?

In a better world, there would be no “I ran some algos on it and now it’s mine” defense.

yangcheng• 2 years ago

latest llama 3.1 is in a different repo, https://github.com/meta-llama/llama-models/blob/main/models/... , but yes, the code is shared. It astonishing that in software 2.0 era, powerful applications like llama has only hundreds of lines of code, and most work hidden in training data. Source code alone is no longer that informative as Software 1.0

twelvechairs• 2 years ago

Open training data would be great too.

If you have open data and open source code you can reproduce the weights

ajxlasA• 2 years ago

Really? I have to check out the training code again. Last time I looked the training and inference code were just example toys that were barely usable.

Has that changed?

danielrhodes• 2 years ago

For models of this size, the code used to train them is going to be very custom to the architecture/cluster they are built on. It would be almost useless to anybody outside of Meta. The dataset would be more a lot more interesting, as it would at the very least show everybody how they got it to behave in certain ways.

mkolodny2• 2 years ago

Llama’s code is open source: https://github.com/meta-llama/llama3/blob/main/llama/model.p...

apsec112• 2 years ago

That's not the training code, just the inference code. The training code, running on thousands of high-end H100 servers, is surely much more complex. They also don't open-source the dataset, or the code they used for data scraping/filtering/etc.

blackeyeblitzar9• 2 years ago

That is just the inference code. Not training code or evaluation code or whatever pre/post processing they do.

mesebrec8• 2 years ago

This is like saying any python program is open source because the python runtime is open source.

Inference code is the runtime; the code that runs the model. Not the model itself.

Flimm4• 2 years ago

No, it's not. The Llama 3 Community License Agreement is not an open source license. Open source licenses need to meet the criteria of the only widely accepted definition of "open source", and that's the one formulated by the OSI [0]. This license has multiple restrictions on use and distribution which make it not open source. I know Facebook keeps calling this stuff open source, maybe in order to get all the good will that open source branding gets you, but that doesn't make it true. It's like a company calling their candy vegan while listing one its ingredients as pork-based gelatin. No matter how many times the company advertises that their product is vegan, it's not, because it doesn't meet the definition of vegan.

[0] https://opensource.org/osd

stale20024• 2 years ago

Ok call it Open Weights then if the dictionary definitions matter so much to you.

The actual point that matters is that these models are available for most people to use for a lot of stuff, and this is way way better than what competitors like OpenAI offer.

the8thbit20• 2 years ago

They don't "[allow] developers to modify its code however they want", which is a critical component of "open source", and one that Meta is clearly trying to leverage in branding around its products. I would like them to start calling these "public weight models", because what they're doing now is muddying the waters so much that "open source" now just means providing an enormous binary and an open source harness to run it in, rather than serving access to the same binary via an API.

input_sh• 2 years ago

Open Source Initiative (kind of a de-facto authority on what's open source and what not) is spending a whole lot of time figuring out what it means for an AI system to be open source. In other words, they're basically trying to come up with a new license because the existing ones can't easily apply.

I believe this is the current draft: https://opensource.org/deepdive/drafts/the-open-source-ai-de...

downWidOutaFite• 2 years ago

OSI made themselves the authority because they hated Richard Stallman and his Free Software movement. It's just marketing.

Zambyte5• 2 years ago

If so, then how can current ML models be open source?

The source of a language model is the text it was trained on. Llama models are not open source (contrary to their claims), they are open weight.

thayne2• 2 years ago

I think it would also include the code used to train it

moffkalast3• 2 years ago

You can find the entire Llama 3.0 pretraining set here: https://huggingface.co/datasets/HuggingFaceFW/fineweb

15T tokens, 45 terrabytes. Seems fairly open source to me.

root_axis3• 2 years ago

No. The text is an asset used by the source to train the model. The source can process arbitrary text. Text is just text, it was written for communication purposes, software (defined by source code) processes that text in a particular way to train a model.

gorgoiler• 2 years ago

One counterpoint is that major publications (eg New York Times) would have you believe that AI is a mildly lossy compression algorithm capable of reconstructing the original source material.

actinium226• 2 years ago

It's not?

_flux2• 2 years ago

I believe it is able to reconstruct parts of the original source material—if the interrogator already knows the original source material to prompt the model appropriately.

halflings• 2 years ago

Training code is only useful to people in academia, and the closest thing to "code you can modify" are open weights.

People are framing this as if it was an open-source hierarchy, with "actual" open-source requiring all training code to be shared. This is not obvious to me, as I'm not asking people that share open-source libraries to also share the tools they used to develop them. I'm also not asking them to share all the design documents/architecture discussion behind this software. It's sufficient that I can take the end result and reshape it in any way I desire.

This is coming from an LLM practitioner that finetunes models for a living; and this constant debate about open-source vs open-weights seems like a huge distraction vs the impact open-sourcing something like Llama has... this is truly a Linux-like moment. (at a much smaller scale of course, for now at least)

kemiller• 2 years ago

I dunno — if an open source project required, say, a proprietary compiler, that would diminish its open source-ness. But I agree it's not totally comparable, since the weights are not particularly analogous to machine code. We probably need a new term. Open Weights.

nothrowaways2• 2 years ago

Weight is the new code.

nomel2• 2 years ago

I think saying it's the new binary is closer to the truth. You can't reproduce it, but you can use it. In this new version, you can even nudge it a bit to do something a little different.

New stuff, so probably not good to force old words, with known meanings, onto new stuff.

GuB-42• 2 years ago

I like the term "open weights". Open source would be the dataset and code that generates these weights.

There is still a lot you can do with weights, like fine tuning, and it is arguably more useful as retraining the entire model would cost millions in compute.

szundi• 2 years ago

Open source = reproducible binaries (weights) by you on your computer, IMO.

Strategy of FB is that they are good to be a user only and fine ruining competitor’s business with good enough free alternatives while collecting awards as saviors of whatever.

j_maffe2• 2 years ago

Not sure what you mean by "they are good to be a user only." Whatever their strategy is, this is great for the community.

ric2b• 2 years ago

If that were the definition then any software you can install on your computer would be open source. It makes open source lose nearly all meaning.

Just say "open weights", not "open source".

rmbyrro4• 2 years ago

If you think about LLMs as a new kind of programming runtime, the matrices are the source.

beloch• 2 years ago

It's no secret that implementing AI usually involves far more investment into training and teaching than actual code. You can know how a neural net or other ML model works. You can have all the code before you. It's still a huge job (and investment) to do anything practical with that. If Meta shares the code their AI runs on with you, you're not going to be able to do much with it unless you make the same investment in gathering data and teaching to train that AI. That would probably require data Meta won't share. You'd effectively need your own Facebook.

If everyone open sources their AI code, Meta can snatch the bits that help them without much fear of helping their direct competitors.

the8thbit20• 2 years ago

I think you're misunderstanding what I'm saying. I don't think its technically feasible for current models to be open source, because there is no source code to open. Yes, there is a harness that runs the model, but the vast, vast amount of instructions are contained in the model weights, which are akin to a compiled binary.

If we make large strides in interpretability we may have something resembling source code, but we're certainly not there yet. I don't think the solution to that problem should be to change the definition of open source and pretend the problem has been solved.

seoulmetro3• 2 years ago

Unfortunately open source really just means an open API these days. The API is heavily intertwined with closed source.

langcss• 2 years ago

Coming up with the words and concepts to describe the models is a challenge.

Does the training data require permission from the copyright holder to use? Are the weights really open source or more like compiled assembly?

shdjkKA• 2 years ago

Of course you are right, I'd put it less carefully: The quoted Linux line is deceptive marketing.

- If we start with the closed training set, that is closed and stolen, so call it Stolen Source.

- What is distributed is a bunch of float arrays. The Llama architecture is published, but not the training or inference code. Without code there is no open source. You can as well call a compiler book open source, because it tells you how to build a compiler.

Pure marketing, but predictably many people follow their corporate overlords and eagerly adopt the co-opted terms.

Reminder again that FB is not releasing this out of altruism, but because they have an existing profitable business model that does not depend on generated chats. They probably do use it internally for tracking and building profiles, but that is the same as using Linux internally, so they release the weights to destroy the competition.

Isn't price dumping an anti trust issue?

bjornsing• 2 years ago

The term “source code” can mean many things. In a legal context it’s often just defined as the preferred format for modification. It can be argued that for artificial neural networks that’s the weights (along with code and preferably training data).

kashyapc3• 2 years ago

I agree; there's a lot of muddiness in the term "open source AI". Earlier this year there was a talk[1] at FOSDEM, titled "Moving a step closer to defining Open Source AI". It is from someone at the Open Source Initiative. The video and slides are available in the link below[1]. From the abstract:

"Finding an agreement on what constitutes Open Source AI is the most important challenge facing the free software (also known as open source) movement. European regulation already started referring to "free and open source AI", large economic actors like Meta are calling their systems "open source" despite the fact that their license contain restrictions on fields-of-use (among other things) and the landscape is evolving so quickly that if we don't keep up, we'll be irrelevant."

[1] https://fosdem.org/2024/schedule/event/fosdem-2024-2805-movi... defining-open-source-ai/

rbits• 2 years ago

You release all the technology and the training data. Everything that was used to create the model, including instructions.

I'm not sure if facebook has done that

roguas• 2 years ago

No, open source means that sources are open, typically for inspection, modification etc. Also here it can be considered the case. Likely in order to claim "true open source", they would have to share dataset? But even this might not be enough for truely open source model? This dataset is nothing but another artifact. So how did they arrive at this dataset, now they have to share pipelines and infra...

.. the thing is, we have not dealt with llm much, it's hard to say what can be considered open source llm just yet, so we use that as metaphore for now

Oras• 2 years ago

This is obviously good news, but __personally__ I feel the open-source models are just trying to catch up with whoever the market leader is, based on some benchmarks.

The actual problem is running these models. Very few companies can afford the hardware to run these models privately. If you run them in the cloud, then I don't see any potential financial gain for any company to fine-tune these huge models just to catch up with OpenAI or Anthropic, when you can probably get a much better deal by fine-tuning the closed-source models.

Also this point:

We need to protect our data. Many organizations handle sensitive data that they need to secure and can’t send to closed models over cloud APIs.

First, it's ironic that Meta is talking about privacy. Second, most companies will run these models in the cloud anyway. You can run OpenAI via Azure Enterprise and Anthropic on AWS Bedrock.

simonw2• 2 years ago

"Very few companies can afford the hardware to run these models privately."

I can run Llama 3 70B on my (64GB RAM M2) laptop. I haven't tried 3.1 yet but I expect to be able to run that 70B model too.

As for the 405B model, the Llama 3.1 announcement says:

To support large-scale production inference for a model at the scale of the 405B, we quantized our models from 16-bit (BF16) to 8-bit (FP8) numerics, effectively lowering the compute requirements needed and allowing the model to run within a single server node.

InDubioProRubio• 2 years ago

CrowdStrike just added "Centralized Company Controlled Software Ecosystem" to every risk data sheet on the planet. Everything futureproof is self-hosted and open source.

mesebrec8• 2 years ago

Note that Meta's models are not open source in any interpretation of the term.

* You can't use them for any purpose. For example, the license prohibits using these models to train other models. * You can't meaningfully modify them given there is almost no information available about the training data, how they were trained, or how the training data was processed.

As such, the model itself is not available under an open source license and the AI does not comply with the "open source AI" definition by OSI.

It's an utter disgrace for Meta to write such a blogpost patting themselves on the back while lying about how open these models are.

ChadNauseam• 2 years ago

you can't meaningfully modify them given there is almost no information available about the training data, how they were trained, or how the training data was processed.

I was under the impression that you could still fine-tune the models or apply your own RLHF on top of them. My understanding is that the training data would mostly be useful for training the model yourself from scratch (possibly after modifying the training data), which would be extremely expensive and out of reach for most people

mesebrec8• 2 years ago

Indeed, fine-tuning is still possible, but you can only go so far with fine-tuning before you need to completely retrain the model.

This is why Silo AI, for example, had to start from scratch to get better support for small European languages.

chasd003• 2 years ago

From what i understand the training data and careful curation of it is the hard part. Everyone wants training data sets to train their own models instead of producing their own.

causal4• 2 years ago

You are definitely allowed to train other models with these models, you just have to give credit in the name, per the license:

If you use the Llama Materials or any outputs or results of the Llama Materials to create, train, fine tune, or otherwise improve an AI model, which is distributed or made available, you shall also include “Llama” at the beginning of any such AI model name.

mesebrec8• 2 years ago

Indeed, this is something they changed in the 3.1 version of the license.

Regardless, the license [1] still has many restrictions, such as the acceptable use policy [2].

[1] https://huggingface.co/meta-llama/Meta-Llama-3.1-8B/blob/mai...

[2] https://llama.meta.com/llama3_1/use-policy

tw043• 2 years ago

In the early days of high-performance computing, the major tech companies of the day each invested heavily in developing their own closed source versions of Unix.

Because they sold the resultant code and systems built on it for money... this is the gold miner saying that all shovels and jeans should be free.

Am I happy Facebook open sources some of their code? Sure, I think it's good for everyone. Do I think they're talking out of both sides of their mouth? Absolutely.

Let me know when Facebook opens up the entirety of their Ad and Tracking platforms and we can start talking about how it's silly for companies to keep software closed.

I can say with 100% confidence if Facebook were selling their AI advances instead of selling the output it produces, they wouldn't be advocating for everyone else to open source their stacks.

JumpCrisscross21• 2 years ago

if Facebook were selling their AI advances instead of selling the output it produces, they wouldn't be advocating for everyone else to open source their stack

You're acting as if commoditizing one's complements is either new or reprehensible [1].

[1] https://gwern.net/complement

tw043• 2 years ago

You're acting as if commoditizing one's complements is either new or reprehensible [1].

I'm acting as if calling on other companies to open source their core product, just because it's a complement for you, and acting as if it's for the benefit of mankind is disingenuous, which it is.

rvnx3• 2 years ago

The source-code to Ad tracking platform is useless to users.

At the end, it's actually Facebook doing the right thing (though they are known for being evil).

It's a bit of an irony.

The supposedly "good" and "open" people like Google or OpenAI, haven't given their model weights.

A bit like Microsoft became the company that actually supports the whole open-source ecosystem with GitHub.

tw043• 2 years ago

The source-code to Ad tracking platform is useless to users.

It's absolutely not useless for developers looking to build a competing project.

The supposedly "good" and "open" people like Google or OpenAI, haven't given their model weights.

Because they're monetizing it... the only reason Facebook is giving it away is because it's a complement to their core product of selling ads. If they were monetizing it, it would be closed source. Just like their Ads platform...

abetusk• 2 years ago

Another case of "open-washing". Llama is not available open source, under the common definition of open source, as the license doesn't allow for commercial re-use by default [0].

They provide their model, with weights and code, as "source available" and it looks like they allow for commercial use until a 700M monthly subscriber cap is surpassed. They also don't allow you to train other AI models with their model:

""" ... v. You will not use the Llama Materials or any output or results of the Llama Materials to improve any other large language model (excluding Meta Llama 3 or derivative works thereof). ... """

[0] https://github.com/meta-llama/llama3/blob/main/LICENSE

sillysaurusx• 2 years ago

They cannot legally enforce this, because they don’t have the rights to the content they trained it on. Whoever’s willing to fund that court battle would likely win.

There’s a legal precedent that says hard work alone isn’t enough to guarantee copyright, i.e. it doesn’t matter that it took millions of dollars to train.

whimsicalism13• 2 years ago

i think these clauses are unenforceable. it's telling that OAI hasn't tried a similar suit despite multiple extremely well-known cases of competitors training on OAI outputs

nuz• 2 years ago

Everyone complaining about not having data access: Remember that without meta you would have openai and anthropic and that's it. I'm really thankful they're releasing this, and the reason they can't release the data is obvious.

mesebrec8• 2 years ago

Without Meta, you would still have Mistral, Silo AI, and the many other companies and labs producing much more open models with similar performance.

Invictus0• 2 years ago

The irony of this letter being written by Mark Zuckerburg at Meta, while OpenAI continues to be anything but open, is richer than anyone could have imagined.

1024core• 2 years ago

"open source AI" ... "open" ... "open" ....

And you can't even try it without an FB/IG account.

Zuck will never change.

causal4• 2 years ago

I think you can use an HF account as well https://huggingface.co/meta-llama

Gracana• 2 years ago

You can also wait a bit for someone to upload quantized variants, finetunes, etc, and download those. FWIW I'm not making a claim about the legality of that, just saying it's an easy way around needing to sign the agreement.

CamperBob27• 2 years ago

It doesn't require an account. You do have to fill in your name and email (and birthdate, although it seems to accept whatever you feed it.)

mvkel3• 2 years ago

It's a real shame that we're still calling Llama "open source" when at best it's "open weights."

Not that anyone would go buy 100,000 H100s to train their own Llama, but words matter. Definitions matter.

sidcool2• 2 years ago

Honest question. As far as LLMs are concerned, isn't open weights same as open source?

mesebrec8• 2 years ago

Open source requires, at the very least, that you can use it for any purpose. This is not the case with Llama.

The Llama license has a lot of restrictions, based on user base size, type of use, etc.

For example you're not allowed to use Llama to train or improve other models.

But it goes much further than that. The government of India can't use Llama because they're too large. Sex workers are not allowed to use Llama due to the acceptable use policy of the license. Then there is also the vague language probibiting discrimination, racism etc.. good luck getting something like that approved by your legal team.

aloe_falsa• 2 years ago

GPL defines the “source code” of a work as the preferred form of the work for making modifications to it. If Meta released a petabyte of raw training data, would that really be easier to extend and adapt (as opposed to fine-tuning the weights)?

paulhilbert• 2 years ago

No, I would argue that from the three main ingredients - training data, model source code and weights - weights are the furthest away from something akin to source code.

They're more like obfuscated binaries. When it comes to fine-tuning only however things shift a little bit, yes.

blackeyeblitzar9• 2 years ago

No open weights are the output of a proprietary and secretive process of training. It’s like sharing a pre compiled application instead of what you need to reproduce the compiled application.

AI2’s OLMo is an example of what open source actually looks like for LLMs:

https://blog.allenai.org/hello-olmo-a-truly-open-llm-43f7e73...

lolinder3• 2 years ago

Source versus weights seems like a really pedantic distinction to make. As you say, the training code and training data would be worthless to anyone who doesn't have compute on the level that Meta does. Arguably, the weights are source code interpreted by an inference engine, and realistically it's the weights that someone is going to want to modify through fine-tuning, not the original training code and data.

The far more important distinction is "open" versus "not open", and I disagree that we should cede that distinction while trying to fight for "source". The Llama license is restrictive in a number of ways (it incorporates an entire acceptable use policy) that make it most definitely not "open" in the customary sense.

mvkel3• 2 years ago

training code and training data would be worthless to anyone who doesn't have compute on the level that Meta does

I don't fully agree.

Isn't that like saying *nix being open source is worthless unless you're planning to ship your own Linux distro?

Knowing how the sausage is made is important if you're an animal rights activist.

JamesBarney• 2 years ago

https://llama.meta.com/llama3_1/use-policy/

The acceptable use policy is seems fine. Don't use it to break the law, solicit sex, kill people, or lie.

frabcus10• 2 years ago

Meta could change the license of future releases of Llama and kill your business built on it.

If the training data was openly available, even if you can't afford to res train a new version, a competitor like Amazon could do it for you

rybosworld• 2 years ago

Huge companies like facebook will often argue for solutions that on the surface, seem to be in the public interest.

But I have strong doubts they (or any other company) actually believe what they are saying.

Here is the reality:

- Facebook is spending untold billions on GPU hardware.

- Facebook is arguing in favor of open sourcing the models, that they spent billions of dollars to generate, for free...?

It follows that companies with much smaller resources (money) will not be able to match what Facebook is doing. Seems like an attempt to kill off the competition (specifically, smaller organizations) before they can take root.

Salgat• 2 years ago

The reason for Meta making their model open source is rather simple: They receive an unimaginable amount of free labor, and their license only excludes their major competitors to ensure mass adoption without benefiting their competition (Microsoft, Google, Alibaba, etc). Public interest, philanthropy, etc are just nice little marketing bonuses as far as they're concerned (otherwise they wouldn't be including this licensing restriction).

noiseinvacuum3• 2 years ago

All correct, Meta does obviously benefit.

It's helpful to also look at what do the developers and companies (everyone outside of top 5/10 big tech companies) get out of this. They get open access to weights of SOTA LLM models that take billions of dollars to train and 10s of billions a year to run the AI labs that make these. They get the freedom to fine tune them, to distill them, and to host them on their own hardware in whatever way works best for their products and services.

frabcus10• 2 years ago

Meta haven't made an open source model. They have released a binary with a proprietary but relatively liberal license. Binaries are not source and their license isn't free.

mattnewton2• 2 years ago

I actually think this is one of the rare times where the small guys interests are aligned with Meta. Meta is scared of a world where they are locked out of LLM platforms, one where OpenAI gets to dictate rules around their use of the platform much like Apple and Google dictates rules around advertiser data and monetization on their mobile platforms. Small developers should be scared of a world where the only competitive LLMs are owned by those players too.

Through this lense, Meta’s actions make more sense to me. Why invest billions in VR/AR? The answer is simple, don’t get locked out of the next platform, maybe you can own the next one. Why invest in LLMs? Again, don’t get locked out. Google and OpenAi/Microsoft are far larger and ahead of Meta right now and Meta genuinely believes the best way to make sure they have an LLM they control is to make everyone else have an LLM they can control. That way community efforts are unified around their standard.

mupuff12343• 2 years ago

Sure, but don't you think the "not getting locked out" is just the pre-requisite for their eventual goal of locking everyone else out?

myaccountonhn• 2 years ago

I actually think this is one of the rare times where the small guys interests are aligned with Meta

Small guys are the ones being screwed over by AI companies and having their text/art/code stolen without any attribution or adherence to license. I don’t think Meta is on their side at all

ketzo• 2 years ago

Meta is, fundamentally, a user-generated-content distribution company.

Meta wants to make sure they commoditize their complements: they don’t want a world where OpenAI captures all the value of content generation, they want the cost of producing the best content to be as close to free as possible.

chasd003• 2 years ago

i was thinking along the same. A lot of content generated by LLMs is going to end up on Facebook or Instagram. The easier it is to create AI generated content the more content ends up on those applications.

Nesco2• 2 years ago

Especially because genAI is a copyright laundering system. You can train it on copyrighted material and none of the content generated with it are copyright-able, which is perfect for social apps

KaiserPro2• 2 years ago

The model it's self isn't actually that valuable to facebook. The thing that's important is the dataset, the infrastructure and the people to make the models.

There is still, just about, a strong ethos( especially in the research teams) to chuck loads of stuff over the wall into opensource. (pytorch, detectron, SAM, aria etc)

but its seen internally as a two part strategy:

1) strong recruitment tool (come work with us, we've done cool things, and you'll be able to write papers)

2) seeding the research community with a common toolset.

jorblumesea• 2 years ago

Cynically I think this position is largely due to how they can undercut OpenAI's moat.

wayeq• 2 years ago

It's not cynical, it's just an awareness that public companies have a fiduciary duty to their shareholders.

cs7022• 2 years ago

We’re releasing Llama 3.1 405B, the first frontier-level open source AI model, as well as new and improved Llama 3.1 70B and 8B models.

Bravo! While I don't agree with Zuck's views and actions on many fronts, on this occasion I think he and the AI folks at Meta deserve our praise and gratitude. With this release, they have brought the cost of pretraining a frontier 400B+ parameter model to ZERO for pretty much everyone -- well, everyone except Meta's key competitors.[a] THANK YOU ZUCK.

Meanwhile, the business-minded people at Meta surely won't mind if the release of these frontier models to the public happens to completely mess up the AI plans of competitors like OpenAI/Microsoft, Google, Anthropic, etc. Come to think of it, the negative impact on such competitors was likely a key motivation for releasing the new models.

---

[a] The license is not open to the handful of companies worldwide which have more than 700M users.

swyx2• 2 years ago

the AI folks at Meta deserve our praise and gratitude

We interviewed Thomas who led Llama 2 and 3 post training here in case you want to hear from someone closer to the ground on the models https://www.latent.space/p/llama-3

throwaway_2494• 2 years ago

We’re releasing Llama 3.1 405B

Is it possible to run this with ollama?

jessechin• 2 years ago

Sure, if you have a H100 cluster. If you quant it to int4 you might get away with using only 4 H100 GPUs!

vorticalbox• 2 years ago

If you have the ram for it.

Ollama will offload as many layers as it can to the gpu then the rest will run on the cpu/ram.

Havoc7• 2 years ago

If you want your first token around tomorrow lunch sure

tambourine_man2• 2 years ago

Praising is good. Gratitude is a bit much. They got this big by selling user generated content and private info to the highest bidder. Often through questionable means.

Also, the underdog always touts Open Source and standards, so it’s good to remain skeptical when/if tables turn.

sheepscreek2• 2 years ago

All said and done, it is a very expensive and balsy way to undercut competitors. They’ve spent > $5B on hardware alone, much of which will depreciate in value quickly.

Pretty sure the only reason Meta’s managed to do this is because of Zuck’s iron grip on the board (majority voting rights). This is great for Open Source and regular people though!

ricardo81• 2 years ago

selling user generated content and private info to the highest bidder

Was always their modus operandi, surely. How else would they have survived.

Thanks for returning everyone else;s content and never mind all the content stealing your platform did.

jart2• 2 years ago

I'm perfectly happy with them draining the life essence out of the people crazy enough to still use Facebook, if they're funneling the profits into advancing human progress with AI. It's an Alfred Nobel kind of thing to do.

germinalphrase• 2 years ago

"Come to think of it, the negative impact on such competitors was likely a key motivation for releasing the new models."

"Commoditize Your Complement" is often cited here: https://gwern.net/complement

tintor3• 2 years ago

they have brought the cost of pretraining a frontier 400B+ parameter model to ZERO

It is still far from zero.

cs7022• 2 years ago

If the model is already pretrained, there's no need to pretrain it, so the cost of pretraining is zero.

pwdisswordfishd• 2 years ago

Makes me wonder why he's really doing this. Zuckerberg being Zuckerberg, it can't be out of any genuine sense of altruism. Probably just wants to crush all competitors before he monetizes the next generation of Meta AI.

spiralk2• 2 years ago

Its certainly not altruism. Given that Facebook/Meta owns the largest user data collection systems, any advancement in AI ultimately strengthens their business model (which is still mostly collecting private user data, amassing large user datasets, and selling targeting ads).

There is a demo video that shows a user wearing a Quest VR headset and asks the AI "what do you see" and it interprets everything around it. Then, "what goes well with these shorts"... You can see where this is going. Wearing headsets with AIs monitoring everything the users see and collecting even more data is becoming normalized. Imagine the private data harvesting capabilities of the internet but anywhere in the physical world. People need not even choose to wear a Meta headset, simply passing a user with a Meta headset in public will be enough to have private data collected. This will be the inevitable result of vision models improvements integrated into mobile VR/AR headsets.

phyrex• 2 years ago

You can always listen to the investor calls for the capitalist point of view. In short, attracting talent, building the ecosystem, and making it really easy for users to make stuff they want to share on Meta's social networks

bun_at_work3• 2 years ago

I really think the value of this for Meta is content generation. More open models (especially state of the art) means more content is being generated, and more content is being shared on Meta platforms, so there is more advertising revenue for Meta.

chasd003• 2 years ago

All the content generated by llms (good or bad) is going to end up back in Facebook/Instagram and other social media sites. This enables Meta to show growth and therefore demand a higher stock price. So it makes sense to get content generation tools out there as widely as possible.

zmmmmm5• 2 years ago

He's not even pretending it's altruism. Literally about 1/3 of the entire post is the section titled "Why Open Source AI Is Good for Meta". I find it really weird that there are whole debates in threads here about whether it's altruistic when Zuckerberg isn't making that claim in the first place.

cageface• 2 years ago

He addresses this pretty clearly in the post. They don't want to be beholden to other companies to build the products they want to build. Their experience being under Apple's thumb on mobile strongly shaped this point of view.

GreenWatermelon4• 2 years ago

Zuckerberg didn't really say anything about altruism. The point he was making is an explicit "I believe open models are best for our business"

He was clear in that one of their motivations is avoiding vendor lockin. He doesn't want Meta to be under the control of their competitors or other AI providers.

He also recognizes the value brought to his company by open sourcing products. Just look at React, PyTorch, and GraphQL. All industry standards, and all brought tremendous value to Facebook.

troupo4• 2 years ago

There's nothing open source about it.

It's a proprietary dump of data you can't replicate or verify.

What were the sources? What datasets it was trained on? What are the training parameters? And so on and so on

advael7• 2 years ago

Look, absolutely zero people in the world should trust any tech company when they say they care about or will keep commitments to the open-source ecosystem in any capacity. Nevertheless, it is occasionally strategic for them to do so, and there can be ancillary benefits for said ecosystem in those moments where this is the best play for them to harm their competitors

For now, Meta seems to release Llama models in ways that don't significantly lock people into their infrastructure. If that ever stops being the case, you should fork rather than trust their judgment. I say this knowing full well that most of the internet is on AWS or GCP, most brick and mortar businesses use Windows, and carrying a proprietary smartphone is essentially required to participate in many aspects of the modern economy. All of this is a mistake. You can't resist all lock-in. The players involved effectively run the world. You should still try where you can, and we should still be happy when tech companies either slip up or make the momentary strategic decision to make this easier

ori_b2• 2 years ago

If that ever stops being the case, you should fork rather than trust their judgment.

Fork what? The secret sauce is in the training data and infrastructure. I don't think either of those is currently open.

ladzoppelin• 2 years ago

Is forking really possible with an LLM or one the size of future Lama versions, have they even released the weights and everything? Maybe I am just negative about it because I feel Meta is the worst company ever invented and feel this will hurt society in the long run just like Facebook.

holoduke• 2 years ago

In tech you can trust the underdogs. Once they turn into dominant players they turn evil. 99% of the cases.

sandworm101• 2 years ago

> Bravo! While I don't agree with Zuck's views and actions on many fronts, on this occasion I think he and the AI folks at Meta deserve our praise and gratitude.

Nope. Not one bit. Supporting F/OSS when it suits you in one area and then being totally dismissive of it in every other area should not be lauded. How about open sourcing some of FB's VR efforts?

zmmmmm5• 2 years ago

They open sourced Gear VR. The are a stark contrast to other players in terms of how they have built everything on open standards (OpenXR, WebXR, etc), and they have just opened their platform by allowing third parties to build and customise it to make their own commercial offerings. Not open source, but a quite a contrast to every other player in that industry so far.

y04nn• 2 years ago

Don't be fooled, it is a "embrace extend extinguish" strategy. Once they have enough usage and be the default standard they will start to find any possible ways to make you pay.

war3215• 2 years ago

Hasn't really happened with PyTorch or any of their other open sourced releases tbh.

GreenWatermelon4• 2 years ago

Credits where due: Facebook didn't do that with React or PyTorch. Meta will reap benefit for sure, but they don't seem to be betting on selling the model itself, rather they will benefit from being at the forefront of a new ecosystem.

tyler-jn• 2 years ago

So far, it seems like this release has done ~nothing to the stock price for GOOGL/MSFT, which we all know has been propped up largely on the basis of their AI plans. So it's probably premature to say that this has messed it up for them.

userabchn• 2 years ago

Interview with Mark Zuckerberg released today: https://www.bloomberg.com/news/videos/2024-07-23/mark-zucker...

FinchNova12• 2 years ago

Our adversaries are great at espionage, stealing models that fit on a thumb drive is relatively easy, and most tech companies are far from operating in a way that would make this more difficult.

Mostly unrelated to the correctness of the article, but this feels like a bad argument. AFAIK, Anthropic/OpenAI/Google are not having issues with their weights being leaked (are they?). Why is it that Meta's model weights are?

meowface• 2 years ago

AFAIK, Anthropic/OpenAI/Google are not having issues with their weights being leaked. Why is it that Meta's model weights are?

The main threat actors there would be powerful nation-states, in which case they'd be unlikely to leak what they've taken.

It is a bad argument though, because one day possession of AI models (and associated resources) might confer great and dangerous power, and we can't just throw up our hands and say "welp, no point trying to protect this, might as well let everyone have it". I don't think that'll happen anytime soon, but I am personally somewhat in the AI doomer camp.

whimsicalism13• 2 years ago

We have no way of knowing whether nation-state level actors have access to those weights.

skybrian3• 2 years ago

I think it’s hard to say. We simply don’t know much from the outside. Microsoft has had some pretty bad security lapses, for example around guarding access to Windows source code. I don’t think we’ve seen a bad security break-in at Google in quite a few years? It would surprise me if Anthropic and OpenAI had good security since they’re pretty new, and fast-growing startups have a lot of organizational challenges.

It seems safe to assume that not all the companies doing leading-edge LLM’s have good security and that the industry as a whole isn’t set up to keep secrets for long. Things aren’t locked down to the level of classified research. And it sounds like Zuckerberg doesn’t want to play the game that way.

At the state level, China has independent AI research efforts and they’re going to figure it out. It’s largely a matter of timing, which could matter a lot.

There’s still an argument to be made against making proliferation too easy. Just because states have powerful weapons doesn’t mean you want them in the hands of people on the street.

dfadsadsf• 2 years ago

We have nationals/citizens of every major US adversary working in those companies with looser security practice than security at local warehouse. Security check before hiring is a joke (mostly checks that resume checks out), laptops can be taken home and internal communication are not segmented on need to know basis. Essentially if China wants weights or source code, it will have hundreds of people to choose from who can provide it.

probablybetter• 2 years ago

I would avoid Facebook and Meta products in general. I do NOT trust them. We have approx. 20 years of their record to go upon.

diggan7• 2 years ago

Today we’re taking the next steps towards open source AI becoming the industry standard. We’re releasing Llama 3.1 405B, the first frontier-level open source AI model,

Why do people keep mislabeling this as Open Source? The whole point of calling something Open Source is that the "magic sauce" of how to build something is publicly available, so I could built it myself if I have the means. But without the training data publicly available, could I train Llama 3.1 if I had the means? No wonder Zuckerberg doesn't start with defining what Open Source actually means, as then the blogpost would have lost all meaning from the get go.

Just call it "Open Model" or something. As it stands right now, the meaning of Open Source is being diluted by all these companies pretending to doing one thing, while actually doing something else.

I initially got very exciting seeing the title and the domain, but hopelessly sad after reading through the article and realizing they're still trying to pass their artifacts off as Open Source projects.

valine• 2 years ago

The codebase to do the training is way less valuable than the weights for the vast majority of people. Releasing the training code would be nice, but it doesn't really help anyone but Meta's direct competitors.

If you want to train on top of Llama there's absolutely nothing stopping you. Plenty of open source tools to do parameter optimization.

diggan7• 2 years ago

Not just the training code but the training data as well, should be under a permissive license, otherwise you cannot call the project itself Open Source, which Facebook does here.

is way less valuable than the weights for the vast majority of people

The same is true for most Open Source projects, most people use the distributed binaries or other artifacts from the projects, and couldn't care less about the code itself. But that doesn't warrant us changing the meaning of Open Source just because companies feel like it's free PR.

If you want to train on top of Llama there's absolutely nothing stopping you.

Sure, but in order for the intent of Open Source to be true for Llama, I should be able to build this project from scratch. Say I have a farm of 100 A100's, could I reproduce the Llama model from scratch today?

vngzs• 2 years ago

Agreed. The Linux kernel source contains everything you need to produce Linux kernel binaries. The llama source does not contain what you need to produce llama models. Facebook is using sleight of hand to garner favor with open model weights.

Open model weights are still commendable, but it's a far cry from open-source (or even libre) software!

elromulous• 2 years ago

100%. With this licensing model, meta gets to reap the benefits of open source (people contributing, social cachet), without any of the real detriment (exposing secret sauce).

hbn• 2 years ago

Is that even something they keep on hand? Or would WANT to keep on hand? I figured they're basically sending a crawler to go nuts reading things and discard the data once they've trained on it.

If that included, e.g. reading all of Github for code, I wouldn't expect them to host an entire separate read-only copy of Github because they trained on it and say "this is part of our open source model"

jdminhbg6• 2 years ago

Why do people keep mislabeling this as Open Source? The whole point of calling something Open Source is that the "magic sauce" of how to build something is publicly available, so I could built it myself if I have the means. But without the training data publicly available, could I train Llama 3.1 if I had the means?

I don't think not releasing the commit history of a project makes it not Open Source, this seems like that to me. What's important is you can download it, run it, modify it, and re-release it. Being able to see how the sausage was made would be interesting, but I don't think Meta have to show their training data any more than they are obligated to release their planning meeting notes for React development.

Edit: I think the restrictions in the license itself are good cause for saying it shouldn't be called Open Source, fwiw.

thenoblesunfish• 2 years ago

You don't need to have the commit history to see "how it works". ML that works well does so in huge part due to the training data used. The leading models today aren't distinguished by the way they're trained, but what they're trained on.

tempfile3• 2 years ago

For the freedom to change to be effective, a user must be given the software in a form they can modify. Can you tweak an LLM once it's built? (I genuinely don't know the answer)

diggan7• 2 years ago

I don't think not releasing the commit history of a project makes it not Open Source,

Right, I'm not talking about the commit history, but rather that anyone (with means) should be able to produce the final artifact themselves, if they want. For weights like this, that requires at least the training script + the training data. Without that, it's very misleading to call the project Open Source, when only the result of the training is released.

What's important is you can download it, run it, modify it, and re-release it

But I literally cannot download the project, build it and run it myself? I can only use the binaries (weights) provided by Meta. No one can modify how the artifact is produced, only modify the already produced artifact.

That's like saying that Slack is Open Source because if I want to, I could patch the binary with a hex editor and add/remove things as I see fit? No one believes Slack should be called Open Source for that.

unraveller• 2 years ago

Open-weights is not open-source, for sure, but I don't mind it being stated as an aspiration goal, the moment it is legally possible to publish a source without shooting themselves in the foot they should do it.

They could release 50% of their best data but that would only stop them from attracting the best talent.

JeremyNT• 2 years ago

Why do people keep mislabeling this as Open Source?

I guess this is a rhetorical question, but this is a press release from Meta itself. It's just a marketing ploy, of course.

blcknight• 2 years ago

InstructLab and the Granite Models from IBM seem the closest to being open source. Certainly more than whatever FB is doing here.

(Disclaimer: I work for an IBM subsidiary but not on any of these products)

hubraumhugo• 2 years ago

The big winners of this: devs and AI startups

- No more vendor lock-in

- Instead of just wrapping proprietary API endpoints, developers can now integrate AI deeply into their products in a very cost-effective and performant way

- Price race to the bottom with near-instant LLM responses at very low prices are on the horizon

As a founder, it feels like a very exciting time to build a startup as your product automatically becomes better, cheaper, and more scalable with every major AI advancement. This leads to a powerful flywheel effect: https://www.kadoa.com/blog/ai-flywheel

danielmarkbruce3• 2 years ago

It creates the opposite of a flywheel effect for you. It creates a leapfrog effect.

boringg5• 2 years ago

AI might cannabalize a lot of first gen AI businesses.

boringg5• 2 years ago

- Price race to the bottom with near-instant LLM responses at very low prices are on the horizon

Maybe a big price war while the market majors fight out for positioning but they still need to make money off their investments so someone is going to have to raise prices at some point and youll be locked into their system if you build on it.

Havoc7• 2 years ago

locked into their system

There are going to be loads of providers for these open models. Openrouter already has 3 providers for the new 405B model within hours.

wavemode2• 2 years ago

they still need to make money off their investments

Depends on how you define this. Most of the top companies don't care as much about making a profit off of AI inference itself, if the existence of the -feature- of AI inference drives more usage and/or sales of their other products (phones, computers, operating systems, etc.)

That's why, for example, Google and Bing searches automatically perform LLM inference at no cost to the user.

choppaface• 2 years ago

Also the opportunity to run on user compute and on private data. That supports a slate of business models that are incompatible with the mainframe approach.

Including adtech models, which are predominantly cloud-based.

drcode• 2 years ago

and Xi Jingping

mav3ri3k• 2 years ago

I am not deep into llms so I ask this. From my understanding, their last model was open source but it was in a way that you can use them but the inner working were "hidden"/not transparent.

With the new model, I am seeing alot of how open source they are and can be build upon. Is it now completely open source or similar to their last models ?

whimsicalism13• 2 years ago

It's intrinsic to transformers that the inner workings are largely inscrutable. This is no different, but it does not mean they cannot be built upon.

Gradient descent works on these models just like the prior ones.

zmmmmm5• 2 years ago

they give you the code and they give you the model it runs, and you can customise and redistribute both. It's all open source in that respect.

What people are complaining about (totally unreasonably in my view) is obviously Meta is not "open sourcing" all the training data, so nobody can retrain the model from scratch themselves. This argument to me is just silly. The whole point of these models is they distil pretraining on massive data sets you wouldn't have access to otherwise. If you insist on them releasing the data set, they will have to cut it down to 0.1% of the size and you will be getting what you had access to already in the first place.

Flimm4• 2 years ago

That's not the only thing that people are complaining about. Even the code is not open source, despite being called open source.

frabcus10• 2 years ago

They could release the code that gathers and curates the data. Give a reproducible system for getting the pre training data. And presumably they own the post training RLHF stuff so could open that.

Without those you're locked in to them in terms of licensing of future versions.

carimura• 2 years ago

Looks like you can already try out Llama-3.1-405b on Groq, although it's timing out. So. Hugged I guess.

TechDebtDevin• 2 years ago

All the big providers should have it up by end of day. They just change their API configs (they're just reselling you AWS Bedrock).

jamiedg2• 2 years ago

405B and the other Llama 3.1 models are working and available on Together AI. https://api.together.ai

Havoc7• 2 years ago

they're just reselling you AWS Bedrock

Meta announced they have 25 providers ready on day 1, so no it's not all AWS.

mensetmanusman• 2 years ago

It’s easy to support open source AI when the code is 1,000 lines and the execution costs $100,000,000 of electricity.

Only the big players can afford to push go, and FB would love to see OpenAI’s code so they can point it to their proprietary user data.

bun_at_work3• 2 years ago

Meta makes their money off advertising, which means they profit from attention.

This means they need content that will grab attention, and creating open source models that allow anyone to create any content on their own becomes good for Meta. The users of the models can post it to their Instagram/FB/Threads account.

Releasing an open model also releases Meta from the burden of having to police the content the model generates, once the open source community fine-tunes the models.

Overall, this move is good business move for Meta - the post doesn't really talk about the true benefit, instead moralizing about open source, but this is a sound business move for Meta.

jklinger410• 2 years ago

This is a great point. Eventually, META will only allow LLAMA generated visual AI content on its platforms. They'll put a little key in the image that clears it with the platform.

Then all other visual AI content will be banned. If that is where legislation is heading.

natural219• 2 years ago

AI moderators too would be an enormous boon if they could get that right.

KaiserPro2• 2 years ago

It would be good, but the cost per moderation is still really high for it to be practical.

noiseinvacuum3• 2 years ago

Creating content with AI will surely be helpful for social media to some extent but I think it's not that important in larger scheme of things, there's already a vast sea of content being created by humans and differentiation is already in recommending the right content to right people at right time.

More important is the products that Meta will be able to make if the industry standardizes on Llama. They would have the front seat in not just with access the latest unreleased models but also settings the direction of progress and next gen LLM optimizes for. If you're Twitter or Snap or TikTok or compete with Meta on the product then good luck in trying to keep up.

apwell233• 2 years ago

I am not sure I follow this.

1. Is there such a thing as 'attention grabbing AI content' ? Most AI content I see is the opposite of 'attention grabbing'. Kindle store is flooded with this garbage and none of it is particularly 'attention grabbing'.

2. Why would creation of such content, even if it was truly attention grabbing, benefit meta in particular ?

3. How would poliferation of AI content lead to more ad spend in the economy. Ad budgets won't increase because of AI content?

To me this is typical Zuckerberg play. Attach metas name to whatever is trendy at the moment like ( now forgotten) metaverse, cryptocoins and bunch of other failed stuff that was trendy for a second. Meta is NOT an Gen AI company ( or a metaverse company, or a cypto company) like he is scamming ( more like colluding) the market to believe. A mere distraction from slowing user growth on ALL of meta apps.

ppl seem to have just forgotten this https://en.wikipedia.org/wiki/Diem_(digital_currency)

bun_at_work3• 2 years ago

Sure - there is plenty of attention grabbing AI content - it doesn't have to grab _your_ attention, and it won't work for everyone. I have seen people engaging with apps that redo a selfie to look like a famous character or put the person in a movie scene, for example.

Every piece of content in any feed (good, bad, or otherwise) benefits the aggregator (Meta, YouTube, whatever), because someone will look at it. Not everything will go viral, but it doesn't matter. Scroll whatever on Twitter, YouTube Shorts, Reddit, etc. Meta has a massive presence in social media, so content being generated is shared there.

The more content of any type leads to more engagement on the platforms where it's being shared. Every Meta feed serves the viewer an ad (for which Meta is paid) every 3 or so posts (pieces of content). It doesn't matter if the user doesn't like 1/5 posts or whatever, the number of ads still goes up.

visarga• 2 years ago

Meta makes their money off advertising, which means they profit from attention. This means they need content that will grab attention

That is why they hopped on the Attention is All You Need train

modo_mario• 2 years ago

I think the biggest part of it is just that they were behind but also betting on it. This allowed them to get a lot of traction, support and be a notable player in the race whilst still retaining some control. Chances are if someone is going to have a frontrow seat monetizing this it's still them.

resters• 2 years ago

This is really good news. Zuck sees the inevitability of it and the dystopian regulatory landscape and decided to go all in.

This also has the important effect of neutralizing the critique of US Government AI regulation because it will democratize "frontier" models and make enforcement nearly impossible. Thank you, Zuck, this is an important and historic move.

It also opens up the market to a lot more entry in the area of "ancillary services to support the effective use of frontier models" (including safety-oriented concerns), which should really be the larger market segment.

passion__desire• 2 years ago

Probably, Yann Lecun is the Lord Varys here. He has Mark's ear and Mark believes in Yann's vision.

war3215• 2 years ago

Unfortunately, there are a number of AI safety people that are still crowing about how AI models need to be locked down, with some of them loudly pivoting to talking about how open source models aid China.

Plus there's still the spectre of SB-1047 hanging around.

amelius• 2 years ago

One of my [Mark Zuckerberg, ed.] formative experiences has been building our services constrained by what Apple will let us build on their platforms. Between the way they tax developers, the arbitrary rules they apply, and all the product innovations they block from shipping, it’s clear that Meta and many other companies would be freed up to build much better services for people if we could build the best versions of our products and competitors were not able to constrain what we could build.

This is hard to disagree with.

glhaynes• 2 years ago

I think it's very easy to disagree with!

If Zuckerberg had his way, mobile device OSes would let Meta ingest microphone and GPS data 24/7 (just like much of the general public already thinks they do because of the effectiveness of the other sorts of tracking they are able to do).

There are certainly legit innovations that haven't shipped because gatekeepers don't allow them. But there've been lots of harmful "innovations" blocked, too.

throwaway1194• 2 years ago

I strongly suspect that what AI will end up doing is push companies and organizations towards open source, they will eventually realize that code is already being shared via AI channels, so why not do it legally with open source?

talldayo5• 2 years ago

they will eventually realize that code is already being shared via AI channels

Private repos are not being reproduced by any modern AI. Their source code is safe, although AI arguably lowers the bar to compete with them.

whimsicalism13• 2 years ago

OpenAI needs to release a new model setting a new capabilities highpoint. This is existential for them now.

ChrisArchitect• 2 years ago

Llama 3.1 Official Launch

https://news.ycombinator.com/item?id=41046540

baceto123• 2 years ago

The value of AI is in the information used to train the models, not the hardware.

m3kw92• 2 years ago

The truth is we need both closed and open source, they both have their discovery path and advantages and disadvantages, there shouldn’t be a system where one is eliminated over the other. They also seem to be driving each other forward via competition.

typpo• 2 years ago

Thanks to Meta for their work on safety, particularly Llama Guard. Llama Guard 3 adds defamation, elections, and code interpreter abuse as detection categories.

Having run many red teams recently as I build out promptfoo's red teaming featureset [0], I've noticed the Llama models punch above their weight in terms of accuracy when it comes to safety. People hate excessive guardrails and Llama seems to thread the needle.

Very bullish on open source.

[0] https://www.promptfoo.dev/docs/red-team/

swyx2• 2 years ago

is there a #2 to llamaguard? Meta seems curiously alone in doing this kind of, lets call it, "practical safety" work

enriquto4• 2 years ago

It's alarming that he refers to llama as if it was open source.

The definition of free software (and open source, for that mater), is well-established. The same definition applies to all programs, whether they are "AI" or not. In any case, if a program was built by training against a dataset, the whole dataset is part of the source code.

Llama is distributed in binary form, and it was built based on a secret dataset. Referring to it as "open source" is not ignorance, it's malice.

Nesco2• 2 years ago

The training data contains most likely insane amounts of copyrighted material. That’s why virtually none of the “open models” come with their training data

enriquto4• 2 years ago

The training data contains most likely insane amounts of copyrighted material.

If that is the case then the weights must inherit all these copyrights. It has been shown (at least in image processing) that you can extract many training images from the weights, almost verbatim. Hiding the training data does not solve this issue.

But regardless of copyright issues, people here are complaining about the malicious use of the term "open source", to signify a completely different thing (more like "open api").

jdminhbg6• 2 years ago

In any case, if a program was built by training against a dataset, the whole dataset is part of the source code.

I'm not sure why I keep seeing this. What is the equivalent of the training data for something like the Linux kernel?

enriquto4• 2 years ago

What is the equivalent of the training data for something like the Linux kernel?

It's the source code.

For the linux kernel:

          compile(sourcecode) = binary

For llama:

          train(data) = weights

indus• 2 years ago

Is there an argument against Open Source AI?

Not the usual nation-state rhetoric, but something that justifies that closed source leads to better user-experience and fewer security and privacy issues.

An ecosystem that benefits vendors, customers, and the makers of close source?

Are there historical analogies other than Microsoft Windows or Apple iPhone / iOS?

kjkjadksj2• 2 years ago

Lets take the iphone. Secured by the industries best security teams I am sure. Closed source, yet teenagers in eastern europe have cracked into it dozens of times making jailbreaks. Every law enforcement agency can crack into it. Closed source is not a security moat, but a trade protection moat.

finolex1• 2 years ago

Replace "Open Source AI" in "is there an argument against xxx" with bioweapons or nuclear missiles. We are obviously not at that stage yet, but it could be a real, non-trivial concern in the near future.

GaggiX• 2 years ago

Llama 3.1 405B is on par with GPT-4o and Claude 3.5 Sonnet, the 70B model is better than GPT 3.5 turbo, incredible.

itissid2• 2 years ago

How are smaller models distilled from large models, I know of LoRA, quantization like technique; but does distilling also mean generating new datasets for conversing with smaller models entirely from the big models for many simpler tasks?

tintor3• 2 years ago

Smaller models can be trained to match log probs of the larger model. Larger model can be used to generate synthethic data for the smaller model.

Havoc7• 2 years ago

I believe these were separately trained not distilled. Though the 405B was used for synth data for the other two

popcorncowboy• 2 years ago

Developers can run inference on Llama 3.1 405B on their own infra at roughly 50% the cost of using closed models like GPT-4o

Does anyone have details on exactly what this means or where/how this metric gets derived?

rohansood15• 2 years ago

I am guessing these are prices on services like AWS Bedrock (their post is down right now).

PlattypusRex• 2 years ago

a big chunk of that is probably the fact that you don't need to pay someone who is trying to make a profit by running inference off-premises.

wesleyyue• 2 years ago

Just added Llama 3.1 405B/70B/8B to https://double.bot (VSCode coding assistant) if anyone would like to try it.

---

Some observations:

* The model is much better at trajectory correcting and putting out a chain of tangential thoughts than other frontier models like Sonnet or GPT-4o. Usually, these models are limited to outputting "one thought", no matter how verbose that thought might be.

* I remember in Dec of 2022 telling famous "tier 1" VCs that frontier models would eventually be like databases: extremely hard to build, but the best ones will eventually be open and win as it's too important to too many large players. I remember the confidence in their ridicule at the time but it seems increasingly more likely that this will be true.

didip• 2 years ago

Is it really open source though? You can't run these models for your company. The license is extremely restrictive and there's NO SOURCE CODE.

jamiedg2• 2 years ago

Looks like it's easy to test out these models now on Together AI - https://api.together.ai

KingOfCoders• 2 years ago

Open Source AI needs to include training data.

fsndz• 2 years ago

Small language models is the path forward https://medium.com/thoughts-on-machine-learning/small-langua...

pja• 2 years ago

“Commoditise your complement” in action!

manishrana2• 2 years ago

rally useful insights

manishrana2• 2 years ago

really useful insights

bufferoverflow2• 2 years ago

Hard disagree. So far every big important model is closed-source. Grok is sort-of the only exception, and it's not even that big compared to the (already old) GPT-4.

I don't see open source being able to compete with the cutting-edge proprietary models. There's just not enough money. GPT-5 will take an estimated $1.2 billion to train. MS and OpenAI are already talking about building a $100 billion training data center.

How can you compete with that if your plan is to give away the training result for free?

sohamgovande• 2 years ago

Where is the $1.2b number from?

bufferoverflow2• 2 years ago

There are a few numbers floating around, $1.2B being the lowest estimate.

HSBC estimates the training cost for GPT-5 between $1.7B and $2.5B.

Vlad Bastion Research estimates $1.25B - 2.25B.

Some people on HN estimate $10B:

https://news.ycombinator.com/item?id=39860293

smusamashah• 2 years ago

Meta's article with more details on the new LLAMA 3.1 https://ai.meta.com/blog/meta-llama-3-1/

6gvONxR4sf7o• 2 years ago

Third, a key difference between Meta and closed model providers is that selling access to AI models isn’t our business model. That means openly releasing Llama doesn’t undercut our revenue, sustainability, or ability to invest in research like it does for closed providers. (This is one reason several closed providers consistently lobby governments against open source.)

The whole thing is interesting, but this part strikes me as potentially anticompetitive reasoning. I wonder what the lines are that they have to avoid crossing here?

phkahler2• 2 years ago

> ...but this part strikes me as potentially anticompetitive reasoning.

"Commoditize your complements" is an accepted strategy. And while pricing below cost to harm competitors is often illegal, the reality is that the marginal cost of software is zero.

Palomides4• 2 years ago

spending a very quantifiable large amount of money to release something your nominal competitors charge for without having your own direct business case for it seems a little much

frabjoused• 2 years ago

Who knew FB would hold OpenAI's original ideals, and OpenAI now holds early FB ideals/integrity.

boringg5• 2 years ago

FB needed to differentiate drastically. FB is at its best creating large data infra.

krmboya• 2 years ago

Mark Zuckerberg was attacked by the media when it suited their tech billionaire villain narrative. Now there's Elon Musk so Zuckerberg gets to be on the good side again

jmward01• 2 years ago

I never thought I would say this but thanks Meta.

*I reserve the right to remove this praise if they abuse this open source model position in the future.

frabcus10• 2 years ago

If it was actually open source with data and the data curation code releases, they wouldn't be able to abuse it the same way. It is open weights, closed training data.

gooob• 2 years ago

why do they keep training on publicly available online data, god dammit? what the fuck. don't they want to make a good LLM? train on the classics, on the essentials reference manuals for different technologies, on history books, medical encyclopedias, journal notes from the top surgeons and engineers, scientific papers of the experiments that back up our fundamental theories. we want quality information, not recent information. we already have plenty of recent information.

mmmore• 2 years ago

I appreciate that Mark Zuckerberg soberly and neutrally talked about some of the risks from advances in AI technology. I agree with others in this thread that this is more accurately called "public weights" instead of open source, and in that vein I noticed some issues in the article.

This is one reason several closed providers consistently lobby governments against open source.

Is this substantially true? I've noticed a tendency of those who support the general arguments in this post to conflate the beliefs of people concerned about AI existential risk, some of whom work at the leading AI labs, with the position of the labs themselves. In most cases I've seen, the AI labs (especially OpenAI) have lobbied against any additional regulation on AI, including with SB1047[1] and the EU AI Act[2]. Can anyone provide an example of this in the context of actual legislation?

On this front, open source should be significantly safer since the systems are more transparent and can be widely scrutinized. Historically, open source software has been more secure for this reason.

This may be true if we could actually understand what was happening in neural networks, or train them to consistently avoid unwanted behaviors. As things are, the public weights are simply inscrutable black boxes, and the existence of jailbreaks and other strange LLM behaviors show that we don't understand how our training processes create models' emergent behaviors. The capabilities of these models and their influence are growing faster than our understand of them, and our ability to steer them to behave precisely how we want, and that will only get harder as the models get more powerful.

At this point, the balance of power will be critical to AI safety. I think it will be better to live in a world where AI is widely deployed so that larger actors can check the power of smaller bad actors.

This paragraph ignores the concept of offense/defense balance. It's much easier to cause a pandemic than to stop one, and cyberattacks, while not as bad as pandemics, seem to also favor the attacker (this one is contingent on how much AI tools can improve our ability to write secure code). At the extreme, it would clearly be bad if everyone had access to a anti-matter weapon large enough to destroy the Earth; at some level of capability, we have to limit the commands an advanced AI will follow from an arbitrary person.

That said, I'm unsure if limiting public weights at this time would be good regulation. They do seem to have some benefits in increasing research around alignment/interpretability, and I don't know if I buy the argument that public weights are significantly more dangerous from a "misaligned ASI" perspective than many competing closed companies. I also don't buy the view of some in the leading labs that we'll likely have "human level" systems by the end of the decade; it seems possible but unlikely. But I worry that Zuckerberg's vision of the future does not adequately guard against downside risks, and is not compatible with the way the technology will actually develop.

[1] https://thebulletin.org/2024/06/california-ai-bill-becomes-a...

[2] https://time.com/6288245/openai-eu-lobbying-ai-act/

btbuildem• 2 years ago

The "open source" part sounds nice, though we all know there's nothing particularly open about the models (or their weights). The barriers to entry remain the same - huge upfront investments to train your own, and steep ongoing costs for "inference".

Is the vision here to treat LLM-based AI as a "public good", akin to a utility provider in a civilized country (taxpayer funded, govt maintained, non-for-profit)?

I think we could arguably call this "open source" when all the infra blueprints, scripts and configs are freely available for anyone to try and duplicate the state-of-the-art (resource and grokking requirements nonwithstanding)

brrrrrm• 2 years ago

check out the paper. it's pretty comprehensive https://ai.meta.com/research/publications/the-llama-3-herd-o...

openrisk• 2 years ago

Open source "AI" is a proxy for democratising and making (much) more widely useful the goodies of high performance computing (HPC).

The HPC domain (data and compute intensive applications that typically need vector, parallel or other such architectures) have been around for the longest time, but confined to academic / government tasks.

LLM's with their famous "matrix multiply" at their very core are basically demolishing an ossified frontier where a few commercial entities (Intel, Microsoft, Apple, Google, Samsung etc) have defined for decades what computing looks like for most people.

Assuming that the genie is out of the bottle, the question is: what is the shape of end-user devices that are optimally designed to use compute intensive open source algorithms? The "AI PC" is already a marketing gimmick, but could it be that Linux desktops and smartphones will suddenly be "ΑΙ natives"?

For sure its a transformational period and the landscape T+10 yrs could be drastically different...

frabcus10• 2 years ago

Unfortunately it is barely more open source than Windows. Llama 3 weights are binary code and while the license is pretty good it isn't open source.

LarsDu88• 2 years ago

Obligatory reminder of why tech companies subsidize open source projects: https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/

avivo• 2 years ago

The FTC also recently put out a statement that is fairly pro-open source: https://www.ftc.gov/policy/advocacy-research/tech-at-ftc/202...

I think it's interesting to think about this question of open source, benefits, risk, and even competition, without all of the baggage that Meta brings.

I agree with the FTC, that the benefits of open-weight models are significant for competition. The challenge is in distinguishing between good competition and bad competition.

Some kind of competition can harm consumers and critical public goods, including democracy itself. For example, competing for people's scarce attention or for their food buying, with increasingly optimized and addictive innovations. Or competition to build the most powerful biological weapons.

Other kinds of competition can massively accelerate valuable innovation.

The FTC must navigate a tricky balance here — leaning into competition that serves consumers and the broader public, while being careful about what kind of competition it is accelerating that could cause significant risk and harm.

It's also obviously not just "big tech" that cares about the risks behind open-weight foundation models. Many people have written about these risks even before it became a subject of major tech investment. (In other words, A16Z's framing is often rather misleading.) There are many non-big tech actors who are very concerned about current and potential negative impacts of open-weight foundation models.

One approach which can provide the best of both worlds, is for cases where there are significant potential risks, to ensure that there is at least some period of time where weights are not provided openly, in order to learn a bit about the potential implications of new models.

Longer-term, there may be a line where models are too risky to share openly, and it may be unclear what that line is. In that case, it's important that we have governance systems for such decisions that are not just profit-driven, and which can help us continue to get the best of all worlds. (Plug: my organization, the AI & Democracy Foundation; https://ai-dem.org/; is working to develop such systems and hiring.)

whimsicalism13• 2 years ago

making food that people want to buy is good actually

i am not down with this concept of the chattering class deciding what are good markets and what are bad, unless it is due to broad-based and obvious moral judgements.

endorphine2• 2 years ago

Except of 90% of the food in the supermarket shelves out there, which is packed in sugar and conservatives.

tpurves2• 2 years ago

405 sounds like a lot of B's! What do you need to practically run or host that yourself?

sumedh3• 2 years ago

You cannot run it locally

tpurves2• 2 years ago

405 is a lot of B's. What does it take to run or host that?

danielmarkbruce3• 2 years ago

quantize to 0 bit. Run on a potato.

Jokes aside ~ 405b x 2 bytes of memory (FP16), so say 810 gigs, maybe 1000 gigs or so required in reality, need maybe 2 aws p5 instances?

dang2• 2 years ago

Related ongoing thread:

Llama 3.1 - https://news.ycombinator.com/item?id=41046540 - July 2024 (114 comments)

littlestymaar• 2 years ago

I love how Zuck decided to play a new game called “commoditize some other billionaire's business to piss him”, I can't wait until this becomes a trend and we get plenty of open source cool stuff.

If he really wants to replicate Linux's success against proprietary Unices, he needs to release Llama with some kind of GPL equivalent, that forces everyone to play the open source game.

Dwedit• 2 years ago

Without the raw data that trained the model, how is it open source?

suyash• 2 years ago

Open source is a welcome step but what we really need is complete decentralisation so people can run their own private AI Models that keep all the data private to them. We need this to happen locally on laptops, mobile phones, smart devices etc. Waiting for when that will become ubiquitous.

frabcus10• 2 years ago

It is open weights not open source. If you can't train it and don't know the training data and can't use it to train your own models, it is a closed model aa a whole. Even if you have the binary weights.

zoogeny• 2 years ago

Totally tangential thought, probably doomed to be lost in the flood of comments on this very interesting announcement.

I was thinking today about Musk, Zuckerberg and Altman. Each claims that the next version of their big LLMs will be the best.

For some reason it reminded me of one apocryphal cause of WW1, which was that the kings of Europe were locked in a kind of ego driven contest. It made me think about the Nation State as a technology. In some sense, the kings were employing the new technology which was clearly going to be the basis for the future political order. And they were pitting their own implementation of this new technology against the other kings.

I feel we are seeing a similar clash of kings playing out. The claims that this is all just business or some larger claim about the good of humanity seem secondary to the ego stakes of the major players. And when it was about who built the biggest rocket, it felt less dangerous.

It breaks my heart just a little bit. I feel sympathy in some sense for the AIs we will create, especially if they do reach the level of AGI. As another tortured analogy, it is like a bunch of competitive parents forcing their children into adversarial relationships to satisfy the parent's ego.

light_triad• 2 years ago

They are positioning themselves as champions of AI open source mostly because they were blindsided by OpenAI, are not in the infra game, and want to commoditize their complements as much as possible.

This is not altruism although it's still great for devs and startups. All FB GPU investments is primarily for new AI products "friends", recommendations and selling ads.

https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/

baby• 2 years ago

Meta does a good thing

HN spends a day figuring out how it’s actually bad

shnock• 2 years ago

It’s not actually bad, OP’s point is that it is not motivated by altruism. An action can be beneficial to the people without that effect being the incentive

satvikpendem• 2 years ago

No one said it was bad. It's just self interested (as companies generally are) and are using that to have a PR spin on the topic. But again, this is what all companies do and nothing about it is bad per se.

m3kw92• 2 years ago

So you think FB did this with zero benefit to themselves? They did open source so people could improve their models and eventually have a paid tier later either from hosting services or other strategies

sensanaty4• 2 years ago

By virtue of it being Meta, it's automatically bad.

If we lived in a sensible world we'd have nuked Meta into a trillion tiny little pieces some time around the Cambridge Analytica bullshit.

war3215• 2 years ago

They've been working on AI for a good bit now. Open source especially is something they've championed since the mid 2010s at least with things like PyTorch, GraphQL, and React. It's not something they've suddenly pivoted to since ChatGPT came in 2022.

kertoip_1• 2 years ago

They are giving it "for free" because:

* they need LLMs that they can control for features on their platforms (Fb/Instagram, but I can see many use cases on VR too)

* they cannot sell it. They have no cloud services to offer.

So they would spend this money anyways, but to compensate some losses they just decided to use it to fix their PR by contenting developers

sterlind2• 2 years ago

They also reap the benefits of AI researchers across the world using Llama as a base. All their research is immediately applicable to their models. It's also likely a strategic decision to reduce the moat OpenAI is building around itself.

I also think LeCunn opposes OpenAI's gatekeeping at a philosophical/political level. He's using his position to strengthen open-source AI. Sure, there's strategic business considerations, but I wouldn't rule out principled motivations too.

sebastiennight7• 2 years ago

I think people massively underestimate how much time/attention span (and ad revenue) will be up for grabs once a platform really nails the "AI friend" concept. And it makes sense for Meta to position themselves for it.

zmmmmm5• 2 years ago

yes ... I remember when online dating was absolutely cringe / weird thing to do. Ten years later and it's the primary way a whole generation seeks a partner.

It will seem incredibly weird today to have an imaginary friend that you treat as a genuine relationship but I genuinely expect this will happen and become a commonplace thing within the next two decades.

Havoc7• 2 years ago

they were blindsided by OpenAI

Given the mountain of GPUs they bought at precisely the right moment I don't think that's entirely accurate

sumedh3• 2 years ago

Given the mountain of GPUs they bought at precisely the right moment I don't think that's entirely accurate

If I remember correctly, FB didnt buy those GPUs because of Open AI, they were going to buy it anyway but Mark said whatever we are buying let's double it.

brigadier132• 2 years ago

Intentions are overrated. Given how many people with good intentions fuck up everything, I'd rather have actual results, even if the intention is self-serving.

istjohn• 2 years ago

AI is not a "complement" of a social network in the way Spolsky defines the term.

A complement is a product that you usually buy together with another product. Gas and cars are complements. Computer hardware is a classic complement of computer operating systems. And babysitters are a complement of dinner at fine restaurants. In a small town, when the local five star restaurant has a two-for-one Valentine’s day special, the local babysitters double their rates. (Actually, the nine-year-olds get roped into early service.)
All else being equal, demand for a product increases when the prices of its complements decrease.

Smart phones ar a complement of Instagram. VR headsets are a complement of the metaverse. AI could be a component of a social network, but it's not a complement.

anthomtb• 2 years ago

My framework for understanding safety is that we need to protect against two categories of harm: unintentional and intentional. Unintentional harm is when an AI system may cause harm even when it was not the intent of those running it to do so. For example, modern AI models may inadvertently give bad health advice. Or, in more futuristic scenarios, some worry that models may unintentionally self-replicate or hyper-optimize goals to the detriment of humanity. Intentional harm is when a bad actor uses an AI model with the goal of causing harm.

Okay then Mark. Replace "modern AI models" with "social media" and repeat this statement with a straight face.

j_m_b• 2 years ago

We need to protect our data.

This is a very important concern in Health Care because of HIPAA compliance. You can't just send your data over the wire to someone's proprietary API. You would at least need to de-identify your data. This can be a tricky task, especially with unstructured text.

xpe• 2 years ago

Zuck needs to get real. They are Open Weights not Open Source.

Sparkyte• 2 years ago

The real path forward is recognizing what AI is good at and what it is bad at. Focus on making what it is good at even better and faster. Open AI will definitely give us that option but it isn't a miracle worker.

My impression is that AI if done correctly will be the new way to build APIs with large data sets and information. It can't write code unless you want to dump billions of dollars into a solution with millions of dollars of operational costs. As it stands it loses context too quickly to do advance human tasks. BUT this is where it is great at assembling data and information. You know what is great at assembling data and information? APIs.

Think of it this way if we can make it faster and it trains on a datalake for a company it could be used to return information faster than a nested micro-service architecture that is just a spiderweb of dependencies.

Because AI loses context simple API requests could actually be more efficient.

Bluescreenbuddy• 2 years ago

This is how we’ve managed security on our social networks – our more robust AI systems identify and stop threats from less sophisticated actors who often use smaller scale AI systems.

So about all the bots and sock puppets on social media..

pjkundert• 2 years ago

Deployment of PKI-signed distributed software systems to use community-provisioned compute, bandwidth and storage at scale is, now quite literally, the future.

We mostly don’t all want or need the hardware to run these AIs ourselves, all the time. But, when we do, we need lots of it for a little while.

This is what Holochain was born to do. We can rent massive capacity when we need it, or earn money renting ours when we don’t.

All running cryptographically trusted software at Internet scale, without the knowledge or authorization of commercial or government “do-gooders”.

Exciting times!

ayakang31415• 2 years ago

Massive props to AI teams at Meta that released this model open source

ceva• 2 years ago

They have earned so much money on all of their users, this is least they can do to give back to the community, if this can be considered that ;)

animanoir• 2 years ago

"Says the Meta Inc".

seydor• 2 years ago

That assumes LLMs are the path to AI, which is increasingly becoming an unpopular opinion

tmsh2• 2 years ago

Software 2.0 is about open licensing.

I.e., the more important thing - the more "free" thing - is the licensing now.

E.g., I play around with different image diffusion models like Stable Diffusion and specific fine-tuned variations for ControlNet or LoRA that I plug into ComfyUI.

But I can't use it at work because of the licensing. I have to use InvokeAI instead of ComfyUI if I want to be careful and only very specific image diffusion models without the latest and greatest fine-tuning. As others have said - the weights themselves are rather inscrutable. So we're building on more abstract shapes now.

But the key open thing is making sure (1) the tools to modify the weights are open and permissive (ComfyUI, related scripts or parts of both the training and deployment) and (2) the underlying weights of the base models and the tools to recreate them have MIT or other generous licensing. As well as the fine-tuned variants for specific tasks.

It's not going to be the naive construction in the future where you take a base model and as company A you produce company A's fine tuned model and you're done.

It's going to be a tree of fine-tuned models as a node-based editor like ComfyUI already shows and that whole tree has to be open if we're to keep the same hacker spirit where anyone can tinker with it and also at some point make money off of it. Or go free software the whole way (i.e., LGPL or equivalent the whole tree of tools).

In that sense unfortunately Llama has a ways to go to be truly open: https://news.ycombinator.com/item?id=36816395

Palmik3• 2 years ago

In the LLM world there are many open source solutions to find tuning, maybe the best one being from Meta: https://github.com/pytorch/torchtune

In terms of inference and interface (since you mentioned comfy) there are many truly open source options such as vLLM (though there isn't a single really performant open source solution for inference yet).

tmsh2• 2 years ago

Thanks! Good to know.

bjt12345• 2 years ago

The LLAMA 3.1 license addresses some of this.

jameson• 2 years ago

It's hard to say Llama is an "open source" when their license states Meta has full control under certain circumstances

https://raw.githubusercontent.com/meta-llama/llama-models/ma...

2. Additional Commercial Terms. If, on the Llama 3.1 version release date, the monthly active users of the products or services made available by or for Licensee, or Licensee’s affiliates, is greater than 700 million monthly active users in the preceding calendar month, you must request a license from Meta, which Meta may grant to you in its sole discretion, and you are not authorized to exercise any of the rights under this Agreement unless or until Meta otherwise expressly grants you such rights.

__loam• 2 years ago

It should be transparently clear that this move was taken by Meta to drive their competitors out of business in a capital intensive space.

apwell233• 2 years ago

not sure how it drives competitors out of business. OpenAI is losing money on queries not on model creation. This opensource model has no impact of their business model of charging users money to run queries.

on a side note OpenAI is losing users on its own. It doesn't need meta to put it out of business.

systemvoltage• 2 years ago

Tbh, it’s incredibly generous.

nailer• 2 years ago

Llama isn't open source. The license is at https://llama.meta.com/llama3/license/ and includes various restrictions on use, which means it falls outside the rules created by the https://opensource.org/osd

war3215• 2 years ago

Even if it's just open weights and not "true" open source, I'll still give Meta the appreciation of being one of the few big AI companies actually committed to open models. In an ecosystem where groups like Anthropic and OpenAI keep hemming and hawing about safety and the necessity of closed AI systems "for our sake", they stand out among the rest.

meowtimemania• 2 years ago

Why would openai/anthropic's approach be more safe? Are people able to remove all the guard rails on the llama models?

alfalfasprout• 2 years ago

They're not safer. The claim is that OpenAI will enforce guard rails and take steps to ensure model outputs and prompts are responsible... but only a fool would take them at their word.

Der_Einzige• 2 years ago

Yes, it's trivial to remove guardrails from any open access model:

https://www.lesswrong.com/posts/jGuXSZgv6qfdhMCuJ/refusal-in...

https://huggingface.co/failspy/Llama-3-70B-Instruct-ablitera...

loceng• 2 years ago

To me it will be most interesting to see who attempts to manipulate the models by stuffing them with content, essentially adding "duplicate" content such as via tautology, in order to make it have added-misallocated weight; which I don't think an AI model will automatically be able to determine, unless it was truly intelligent, instead it would require to be trained by competent humans.

And so the models that have mechanisms for curating and preventing such misapplied weighting, and then the organizations and individuals who accurately create adjustments to the models, will in the end be the winners - where truth has been more honed for.

rednafi• 2 years ago

How’s only sharing the binary artifact is open source? There’s the data aspect of things that they can’t share because of licensing and the code itself isn’t accessible.

Palmik3• 2 years ago

It's much better than sharing a binary artifact of regular software, since the weights can be and are easily and frequently modified by fine tuning the model. This means you can modify the "binary artifact" to your needs, similar to how you might change the code of open source software to add features etc.

bigmattystyles• 2 years ago

Am I too skeptical or is this the approach taken when they've decided they can't make proprietary work for them.

nothrowaways2• 2 years ago

I love llamas but how fb can profit from leading this effort is not clear to me.

AMICABoard• 2 years ago

Okay if anyone wants to try Llama 3.1 inference on CPU, try this: https://github.com/trholding/llama2.c (L2E)

It's a bit buggy but it is fun.

Disclaimer: I am the author of L2E

frays• 2 years ago

No matter how "open source" they actually will be, I'm glad this exists as a competitor to Gemini and ChatGPT to help push innovation from a different angle.

Can't wait to see how the landscape will look in 2027 and beyond.

rbren• 2 years ago

If you're interested in getting Llama 3.1 to build software, check out https://github.com/OpenDevin/OpenDevin

sebastiennight7• 2 years ago

I've summarized this entire thread in 4 lines (didn't even use AI for it!)

Step 1. Chick-Fil-A releases a grass-fed beef burger to spite other fast-food joints, calls it "the vegan burger"

Step 2. A couple of outraged vegans show up in the comments, pointing out that beef, even grass-fed beef, isn't vegan

Step 3. Fast food enthusiasts push back: it's unreasonable to want companies to abide by this restrictive definition of "vegan". Clearly this burger is a gamechanger and the definition needs to adapt to the times.

Step 4. Goto Step 2 in an infinite loop

nathansherburn• 2 years ago

Open source software is one of our best and most passionately loved inventions. It'd be much easier to have a nuanced discussion about "open weights" but I don't think that's in Facebook's interest.

llm_trw6• 2 years ago

More like vegetarians show up claiming to be vegans, then vegans show up and explain why eating animal products is still wrong.

That's the difference between open source and free software.

dogcomplex3• 2 years ago

Yeah the moral step up from the status quo is still laudable. Open weights are still much improved over the closed creepy spy agency clusterfucks that OpenAI/Microsoft/Google/Apple are bringing to the table.

dilliwal• 2 years ago

On point, and pretty good analogy

BaculumMeumEst3• 2 years ago

What is the best of the llama 3.1 models that I can fine-tune with a macbook m3 max w/ 96GB of ram?

lostmsu3• 2 years ago

None unless you are prepared to spend 5+ years. Macs just don't have flops.

BaculumMeumEst3• 2 years ago

Really? You can't even fine-tune a quantized 8B model on such a machine? That's a bummer.

fimdomeio• 2 years ago

This is very amusing:

- We need to control our own destiny and not get locked into a closed vendor. - We need to protect our data. - We want to invest in the ecosystem that’s going to be the standard for the long term.

Thank you Meta for being the bright light of ethical guidance for us all.

dev1ycan2• 2 years ago

Reality: they've realize gpt 4 is a wall, they can't keep pouring trillions of dollars into it for no improvement or little at all, so now they want to put it to the open source until someone figures out the next step then they'll take it behind closed doors again.

I hate how the moment it's too late will be, by design, closed doors.

Ukv• 2 years ago

Reality: they've realize gpt 4 is a wall, they can't keep pouring trillions of dollars into it for no improvement or little at all, so now they want to put it to the open source [...]

This is Meta (LLaMA, which has had available weights for a while), not OpenAI (GPT).

dev1ycan2• 2 years ago

How does that change anything about my comment? what does "available weights" change about the system being closed source? additionally, they have the developers, the second they figure out a way to achieve AGI or something close to it they'll take it closed source, this is just outsourcing maintenance and small tweaks.

smashah• 2 years ago

Actually open source Whatsapp is the way forward.

petetnt• 2 years ago

Great, now release the datasets used for training your AI so everyone can get compensated accordingly and ask that your competition follow suit.

twelve40• 2 years ago

It'll be interesting to come back here in a couple of years and see what's left. What do they even do anymore? They have Facebook, which hasn't visibly changed in a decade. They have Instagram, which feels a bit sleeker but also remained more or less the same. and Whatsapp. Ad network that runs on top of those services and floods them with trash. Bunch of stuff that doesn't seem to exist anymore - Libra, the grandiose multibillion dollar Legless VR, etc.

But they still have 70 thousand people (a small country) doing _something_. What are they doing? Updating Facebook UI? Not really, the UI hasn't been updated, and you don't need 70 thousand people to do that. Stuff like React and Llama? Good, I guess, we'll see how they make use of Llama in a couple of years. Spellcheck for posts maybe?

therealdrag0• 2 years ago

And still making 135B dollars in revenue, or 2M per employee. I don’t know what they do either lol, but I don’t mind that revenue supporting jobs.

benreesman• 2 years ago

In general I look back on my time at FB with mixed feelings, I’m pretty skeptical that modern social media is a force for good and I was there early enough to have moved the needle.

But this is really positive stuff and it’s nice to view my time there through the lens of such a change for the better.

Keep up the good work on this folks.

Time to start thinking about opening up a little on the training data.

bentice• 2 years ago

Ironically, this benefits Apple so much.

netsec_burn• 2 years ago

How? They are prohibited from using it in the license.

ysofunny• 2 years ago

or else is not even scientific

elecush• 2 years ago

Ok one notable difference: did the linux researchers of yore warn about adversarial giants getting this tech? Or is this unique to the current moment? That for me is the largest question when considering the logical progression on "linux open is better therefore ai open is better".

Spivak3• 2 years ago

We can't open source Linux because bad people might run servers?

Can you imagine the disinformation they could spread with those? With enough of them you could have a massively global site made entirely for spreading it. God what if such a thing got into the hands of an egocentric billionaire?

endorphine2• 2 years ago

An operating system is not "generative", hence it's not a force multiplier.

turingbook• 2 years ago

Open-weights models are not really open source.

scotty79• 2 years ago

It's more like a freeware than open source. You can launch it on your hardware and use it but how it was created is mostly not published.

Still huge props to them for doing what they do.

casebash• 2 years ago

I expect this to end up having been one of the worst timed blog posts in history. Open source AI has mostly been good for the world up until now, but we're getting to the point where we're about to find out why open-sourcing sufficiently bad models is a terrible idea.

tananaev• 2 years ago

I don't think weights is the source. Data is the source. But still better than nothing.

cratermoon• 2 years ago

You first, Zuck.

ofou• 2 years ago

                Llama 3 Training System
                 Total: 19.2 exaFLOPS
                         |
            +-------------+-------------+
            |                           |
      Cluster 1               Cluster 2
    9.6 exaFLOPS             9.6 exaFLOPS
           |                       |
    +------+------+         +------+------+
    |             |         |             |
 12K GPUs      12K GPUs  12K GPUs      12K GPUs
    |             |         |             |
  [####]       [####]     [####]       [####]
  400+          400+      400+          400+
 TFLOPS/GPU   TFLOPS/GPU TFLOPS/GPU   TFLOPS/GPU

tomjen3• 2 years ago

Re safety: Just release two models, one that has been tuned and one that hasn't.

Claude is supposed to be better, but it is also even more locked down than ChatGPT.

Word will let me write a manifest for a new Nazi party, but Claude is so locked down that it won't find a cartoon in a picture and Gemini... well.

If AIs are not to harm society, they need to enable us to think in new ways.

slowhadoken• 2 years ago

First you’re going to have to write some laws that prevent openwashing and legitimate open source projects from becoming proprietary.

thntk• 2 years ago

When Zuck said spy can easily steal models, I wonder how much of it comes from experiences. I remember they struggled to train OPT not long ago.

On a more serious note, I don't really buy his arguments about safety. First, widespread AI does not reduce unintentional harm but increases it, because the rate of accident is compound. Second, the chance of success for threat actors will increase, because of the asymmetric advantage of gaining access to all open information and hiding their own information. But there is no reverse at this point, I enjoy it while it lasts, AGI will come sooner or later anyway.

Simon_ORourke• 2 years ago

I thoroughly support Meta's open-sourcing of these AI models going forward. However, for a company that absolutely closed down discussions about providing API access to their platform, I'm left wondering what's in it (monetarily) for them by doing this? Is it to simply undercut competition in the space, like some grocery store selling below cost?

yard20103• 2 years ago

Stay assured these guys are working day and night to make our world a darker place

GreenWatermelon4• 2 years ago

Meta open sources its tools: React, GraphQL, PyTorch, and now these new models. Meta seems to be about open sourcing tools, not providing open access to their platforms.

The AI model complements the platform, and their platform is the money maker. They hold the belief that open sourcing their tools benefit their platform on the long run, which is why they're doing it. And in doing so, they aren't under the control of any competitors.

I would say it's more like a grocery store providing free parking, a bus stop, self-checkout, online menu, and free delivery.

msnkarthik• 2 years ago

Interesting discussion! While I agree with Zuckerberg's vision, the comments raise valid concerns. The point about GPU accessibility and cost is crucial. Public clusters are great, but sustainable funding and equitable access are essential to avoid exacerbating existing inequalities. I also resonate with the call for CUDA alternatives. Breaking the dependence on proprietary technology is key for a truly open AI ecosystem. While existing research clusters offer some access, their scope and resources often pale in comparison to what companies like Meta are proposing. We need a multi-pronged approach: open-sourcing models AND investing in accessible infrastructure, diverse hardware options, and sustainable funding models for a truly democratic AI future.

fnordpiglet• 2 years ago

I suspect we are still early in the optimization evolution. The weights are what matter. The ability to run them anywhere might come.

troupo4• 2 years ago

The training datasets and methodology are what matters. None of that is disclosed by anyone

narrator• 2 years ago

3nm chip fabs take years to build. You don't just go to AWS and spin one up. This is the very hard part about AI that breaks a lot of the usual tech assumptions. We have entered a world where suddenly there isn't enough compute, because it's just too damn hard to build capacity and that's different from the past 40 years.

ohthehugemanate• 2 years ago

Has anyone taken apart the llama community license and compared it to validated open source licenses? Red Hat is making a big deal about releasing the Granite LLM released under Apache. Is there a real difference between that and what Llama does?

https://www.redhat.com/en/topics/ai/open-source-llm

rightbyte• 2 years ago

What would be the speed of a querry of running this model from disk on a ordinary PC?

Has anyone tried that?

v3ss0n• 2 years ago

We welcome Mark Zuckerberg's Redemption Arc! Opensource AI Here we go!

rldjbpin• 2 years ago

this is very cool indeed that meta has made available more than they need to in terms of model weights.

however, the "open-source" narrative is being pushed a bit too much like descriptive ML models were called "AI", or applied statistics "data science". with reinforced examples such as this, we start to lose the original meaning of the term.

the current approach of startups or small players "open-sourcing" their platforms and tools as a means to promote network effect works but is harmful in the long run.

you will find examples of terraform and red hat happening, and a very segmented market. if you want the true spirit of open-source, there must be a way to replicate the weights through access to training data and code. whether one could afford millions of GPU hours or not, real innovation would come from remixing the internals, and not just fine-tuning existing stuff.

i understand that this is not realistically going to ever happen, but don't perform deceptive marketing at the same time.

avereveard• 2 years ago

they said, while not releasing the video part of chamleon model

chx• 2 years ago

A total ban on generative AI is the way forward. If the industry refuses to make it safe by self regulating then the regulator must step in and ban it until better, more fine tuned regulation can be made. It is needed to protect our environment, our democracy, our very lives.

yard20103• 2 years ago

Isn't it like banning Christianity? I don't think it can be done

yard20103• 2 years ago

I don't believe in any word coming out from this lizard. He is the most evil villain I know, and I live in the middle east, can you imagine

pandaswo• 2 years ago

Way to go!

tarruda• 2 years ago

From the "Why Open Source AI Is Good for Meta" section, none of the four given reasons seem to justify spending so much money to train these powerful models and give them away for free.

Third, a key difference between Meta and closed model providers is that selling access to AI models isn’t our business model. That means openly releasing Llama doesn’t undercut our revenue, sustainability, or ability to invest in research like it does for closed providers. (This is one reason several closed providers consistently lobby governments against open source.)

Maybe this is a strategic play to hurt other AI companies that depend on this business model?

sensanaty4• 2 years ago

Meanwhile Facebook is flooded with AI-generated slop with hundreds of thousands of other bots interacting with it to boost it to whoever is insane enough to still use that putrid hellhole of a mass-data-harvesting platform.

Dead internet theory is very much happening in real time, and I dread what's about to come since the world has collectively decided to lose their minds with this AI crap. And people on this site are unironically excited about this garbage that is indistinguishable from spam getting more and more popular. What a fucking joke

rocgf• 2 years ago

I agree with the overall sentiment, but it's is not necessarily the case that "the world has collectively decided to lose their minds with this AI crap". You only need a relatively small number of bad actors for this to be the case.

consf4• 2 years ago

I think the situation highlights the need for better regulation

TheAceOfHearts• 2 years ago

I think there's room for a middle-ground. I agree that there's a lot of slop being generated and shared around. Part of it is due to the initial excitement over these AI tools. Part of it is that most AI tools still kinda suck at the moment. In the long term I expect tools to get way better, which I'm hopefully could help enable smaller teams of people than ever before to execute on their creative vision.

Personally I have tons of creative ideas which I think would be interesting and engaging but for which I lack the resources to bring into this world, so I'm hoping that in the long term AI tools can help bridge this gap. I'm really hopeful for a world where people from all over the world can share their creative stories, rather than being mostly limited to a few rich people in Hollywood.

Unfortunately I do expect this to end up being the minority of content, especially as we continue being flooded by increasing amounts of trash. But maybe that's just opening up the opportunity for someone to develop new content curation tools. If anything, even before the rise of AI stuff there were mountains of content, and we saw with the rise of TikTok that a good recommendation algorithm still leaves room for new platforms.

have_faith2• 2 years ago

The feed certainly is, but I suspect most activity left on Facebook is happening in group pages. Groups are the only thing I still log in for as some of them, particularly the local ones, have no other way of taking part. They are also commonly membership by request and actively moderated. If I had the time (and energy) I might put some effort into advocating to moving to something else, but it will be an uphill battle.

consf4• 2 years ago

The challenges of moving to alternative solutions

jrnx• 2 years ago

I'd assume that any platform which get's sufficiently popular will become a bot and AI content target...

baq• 2 years ago

I stopped going on facebook a few years ago and don't miss it; I don't even need messenger as everyone migrated to whatsapp (yes I know, normal people don't want to move to signal, but got quite a few techy friends to migrate). The FB-only groups are indeed a problem, I'm delegating them to my wife.

IF I ever had to go to FB for anything, I'd probably install a wall-removing browser extension. Mobile app is of course out of question.

throwawayfour2• 2 years ago

For me it's the Marketplace. Left FB many years ago only to come back to keep an eye out for used Lego for the kiddos. At least in my region, and for my purposes, Marketplace is miles better than any other competing sites/apps.

throwaway7ahgb2• 2 years ago

Same here, Groups + Marketplace is actually a wealth of information. There are still a few dark patterns but most manageable for a "free" platform.

OPs comments read like we're describing something the SS built (Godwin says hi).

giancarlostoro2• 2 years ago

It is probably a mix of people who got nowhere else to interact with people, and people using Groups. Facebook was where you'd go to talk to all your friends and family, most of my friends have been getting shadowbanned since 2012 ~ so it made me use it less. I got auto striked on my account for making a meme joke about burning a house down due to a giant spider in a video. I appealed, and it got denied. I'm not using a platform that will inadvertently ban me by AI. But the people actually posting to kill others, and actually burn shit down, and bots stay just fine?

Plus I didn't want to risk my employers Facebook App being in limbo if I got banned, so I left Facebook alone, never to return.

Facebook trying to police the world is the only thing keeping me away, if I can use the platform and post meme comments again, maybe I might reconsider, but I doubt it. Reddit is in a similar boat. You can get banned, but all the creepy pedophile comments from decades and recently are still up no problem.

sgu9993• 2 years ago

If I had the time (and energy) I might put some effort into advocating to moving to something else, but it will be an uphill battle.

What are the alternatives for local groups? I've recently seen an increase in the amount of Discourse forums available, which is nice, but I don't think it'd be very appealing to the average cycling or hiking group.

kashyapc3• 2 years ago

I get your frustration of a scorched internet. But I don't think it's all that gloomy. Whether we like it or not, LLMs and some kind of a "down-to-earth AI" is here to stay, once the dust settles. Right now, it feels like everything is burning because we're in the thick of an evolving situation; and the Silicon Valley tech-bros are in a hurry to ride the wave and make a quick buck with their ChatGPT wrapper. (I can't speak of social networks, I don't have any accounts for 10+ years, except for HN.)

    * * *

On "collective losing of minds", you might appreciate this quote from 1841 (!) by Charles MacKay. I quoted it in the past[1] here, but is worth re-posting:

"In reading the history of nations, we find that, like individuals, they have their whims and their peculiarities; their seasons of excitement and recklessness, when they care not what they do. We find that whole communities suddenly fix their minds upon one object, and go mad in its pursuit; that millions of people become simultaneously impressed with one delusion, and run after it, till their attention is caught by some new folly more captivating than the first [...]

"Men, it has been well said, think in herds; it will be seen that they go mad in herds, while they only recover their senses slowly, and one by one."

— from MacKay's book, 'Extraordinary Popular Delusions and the Madness of Crowds'

[1] https://news.ycombinator.com/item?id=25767454

sgu9993• 2 years ago

What a nice quote. What were "millions of people" simultaneously obsessed with around 1841?

LoveMortuus• 2 years ago

In regards to the dead internet hypothesis, the content that you're enjoying today, will still be there tomorrow. What I mean is if you, for example, like Mr.Beast, AI is not going to replace him and the content that he produces. Now, he might use AI to boost the productivity of his company, but the end result is still "making the best video ever" as he's often said.

LauraMedia• 2 years ago

The big problem with this is that content is harder and harder to find. Try to find a non-AI generated reply to a viral post on Twitter, you're looking at having to scroll down 5-6 1080p screens to finally get to some actual stuff people wrote.

The content you're enjoying today still exists, but it's a needle in a haystack of AI spam

elorant• 2 years ago

Every content that exists on the web could now be rewritten and repurposed by LLMs. This could lead into an explosion of web sites that could easily double in size every few years. Good luck indexing all that crap and deciding what is duplicate and what not.

sbeaks• 2 years ago

Are we going to end up with rererecapture where to post you need something on your phone/laptop measuring typing speed and scanning your face with an IR cam? Or a checkmark showing typed out by hand? Wouldn't get rid of ai content but may slow down bots posting it.

sebastiennight7• 2 years ago

I got so tired of having to prove I'm a human, I wrote a bunch of prompts for multimodal LLMs to solve CAPTCHAs for me.

kristopolous2• 2 years ago

They're trajectory is so close to AOL's it's almost implausible. Their cash cow flagship product is widely panned by tech insiders as abusive, manipulative, and toxic but they also place significant financial resources in high quality open source projects out of what can only be described as benevolence and some commitment to the common good.

aranke• 2 years ago

Which open-source projects is AOL known for? A quick Google search isn’t returning much.

csomar• 2 years ago

That's actually the preferred outcome. The open internet noise ratio will be so high that it turns into pure randomness. The traditional old venues (reputed blogs, small communities forums, pay for valued information, pay for your search, etc..) will resurface again. The Popular Web has been in a slow decline, time to kill it.

wildrhythms• 2 years ago

My concern is that these platforms will soon sell Human Created (tm) content back to us.

thierrydamiba2• 2 years ago

Is it really a decline? If people are looking for and consuming the slop, where is the issue?

There is still plenty high quality stuff too if that is what you’re looking for. If you want to roll with the pigs in the shit, who am I to tell you no?

dspillett• 2 years ago

The traditional old venues […] will resurface again.

… to be subsequently drowned out by AI “copies” of themselves, which in turn are used to train more AIs, until we don't have a Dead Internet¹ but a Habsburg Internet.

[1] https://en.wikipedia.org/wiki/Dead_Internet_theory

eightysixfour2• 2 years ago

I am one of those people unironically excited - the social parts of the internet have been dead and filled with bots for a long time. Now people just see it more.

Maybe they’ll go outside.

zwnow• 2 years ago

Bots were easy to detect, now they're almost indistinguishable from human interaction. The death of the internet would be an incredible loss for humanity. You will not be able to trust anything you find online. Nothing is safe from this.

Kiro2• 2 years ago

And I don't understand why people lump all AI together as if a coding assistant is the same thing as AI generated spam and other garbage. I'm pretty sure no-one here is excited about that.

I'm excited about the former since AI has massively improved my productivity as a programmer to a point where I can't imagine going back. Everything is not black or white and people can be excited about one part of something and hate another at the same time.

chr15m• 2 years ago

fear is why

sensanaty4• 2 years ago

Seeing some of the code my colleagues are shitting out with the help of coding "assistants", I would definitely categorize their contributions as spam, and has had nothing but an awful effect on my own time and energy, having to sift through the unfiltered crap. The problem being, of course, that the idiotic C-suite in their infinite wisdom decided to push "use AI assistants" as a KPI, so people are even encouraged to spam PRs with terrible code.

If this is what productivity looks like then I'm proud to be unproductive.

testfrequency• 2 years ago

Hear, hear!

codedokode• 2 years ago

AI should not be open source because it can be used in military applications. It doesn't make sense to give away a technology others might use against you.

ChanderG• 2 years ago

I think all this discussion around Open-source AI is a total distraction from the elephants in the room. Let's list what you need to run/play around with something like Llama:

1. Software: this is all Pytorch/HF, so completely open-source. This is total parity between what corporates have and what the public has.

2. Model weights: Meta and a few other orgs release open models - as opposed to OpenAI's closed models. So, ok, we have something to work with.

3. Data: to actually do anything useful you need tons of data. This is beyond the reach of the ordinary man, setting aside the legality issues.

4. Hardware: GPUs, which are extremely expensive. Not just that, even if you have the top dollars, you have to go stand in a queue and wait for O(months), since mega-corporates have gotten there before you.

For Inference, you need 1,2 and 4. For training (or fine-tuning), you need all of these. With newer and larger models like the latest Llama, 4 is truly beyond the reach of ordinary entities.

This is NOTHING like open-source, where a random guy can edit/recompile/deploy software on a commodity computer. Wrt LLMs, Data/Hardware are in the equation, the playing field is complete stacked. This thread has a bunch of people discussing nuances of 1 and 2, but this bike-shedding only hides the basic point: Control of LLMs are for mega-corps, not for individuals.

fishermanbill3• 2 years ago

But there is an insidiousness to Meta calling their software 'open source'. It feels as if they are riding on the coat tails of the term as if they are being altruistic, when in fact they are being no more altruistic than any large corporation that wants to capture market share via their financial muscle - which I suppose touches on your last point.

fishermanbill3• 2 years ago

Its not open source.

We don't get the data or training code. The small runtime framework is open source but that's of little use as its largely fixed in implementation due to the weights. Yes we can fine tune but that is akin to modifying video games - we can do it but there's only so much you can do within reasonable effort and no one would call most video games 'open source'*.

Its freeware and Meta's strategy is much more akin to the strategy Microsoft used with Internet Explorer to capture the web browser market. No one was saying god bless Microsoft for trying to capture the browser market with I.E. Nothing wrong with Meta's strategy just don't call it open source.

*weights are data and so is the video/audio output of a video game. If we gave away that video game output for free we wouldn't call the video game open source as the myriad freeware games essentially do.

Palmik3• 2 years ago

I don't think these analogies work.

Meta provides open source code to modify the the weights (fine tune the model). In this context, fine-tuning the model is better converted to being able to modify the code of the game.

fishermanbill3• 2 years ago

So do video game developers (provide source code to modify their games) the analogy absolutely works. I can list a huge amount of actually open source software that I can see the source code and data for which is very different from Llama etc.

OriginalMrPink• 2 years ago

Open Source AI is the path forward, but I have hard time believing that Meta should be affiliated with it.

pera• 2 years ago

I wish Meta stopped using the "open source" misnomer for free of charge weights. In the US the FTC already uses the term Open-Weights, and it seems the industry is also adopting this term (e.g. Mistral).

Someone can correct me here but AFAIK we don't even know which datasets are used to train these models, so why should we even use "open" to describe Llama? This is more similar to a freeware than an open-source project.

[1] https://www.ftc.gov/policy/advocacy-research/tech-at-ftc/202...

benrutter• 2 years ago

This is such a good point. The industry is really putting the term "open source" through the ringer at the moment but I don't see any justification for considering the final weight output a "source" anymore than releasing a compiled binary would be open source.

In fairness to Llama, the source code itself (though not the training data) is available to access, although not really under a license that many would consider open source.

zelphirkalt3• 2 years ago

Facebook is one of the great when it comes to twisting words and appropriating terms in ways that benefit Facebook.

largbae• 2 years ago

The All-In podcast predicted this exact strategy for keeping OpenAI and other upstarts from disrupting the existing big tech firms.

By giving away higher and higher quality models, they undermine the potential return on investment for startups who seek money to train their own. Thus investment in foundation model building stops and they control the ecosystem.

giancarlostoro2• 2 years ago

I enjoy that podcast, only really listened to it a few times, but they definitely bring up some interesting topics, the kind I come on HN for.

thierrydamiba2• 2 years ago

I predicted this 7 months ago-can I get a podcast?

https://news.ycombinator.com/item?id=38556771#38559118

zelphirkalt3• 2 years ago

Open data for open algo for open AI is the path forward.

dcist• 2 years ago

So commoditize the complement.

ssahoo2• 2 years ago

Additional Commercial Terms. If, on the Llama 3.1 version release date, the monthly active users of the products or services made available by or for Licensee, or Licensee’s affiliates, is greater than 700 million monthly active users in the preceding calendar month, you must request a license from Meta, which Meta may grant to you in its sole discretion, and you are not authorized to exercise any of the rights under this Agreement unless or until Meta otherwise expressly grants you such rights.

Which open-source has such restrictions and clause?

rmbyrro4• 2 years ago

Which open source costs dedicated usage of 16 thousand H100 over several months?

C'mon folks, they're opening up for free to 99.99% of potential users what cost hundreds of millions of dollars, if not in the ballpark of a billion.

Let's appreciate that, instead of focusing on semantics for a while.

arrosenberg• 2 years ago

I don’t think the largest tech companies in the world have earned that view of benevolence. Its real hard to take altruism seriously when it is coming from Zuckerberg.

birdalbrocum• 2 years ago

Licensing is not a simple semantic problem. It is a legal problem that have strong ramifications, especially things are on their way to standardize. What Facebook is trying to do with their "open source" models is to exhaust possibility of fully open source models to be industry standarts. and create an alternative monopoly to Microsoft/OpenAI. Think of it as if an entity had right to ISO standards, they would be extremely rich. Eventually researchers will release pretty advance ML models that are fully open source(from dataset to training code) and Facebook is trying to block them even before start to prevent of the possibility of this models to be standard. This is a complementary tactic of the industry to closed source rivals and should not be understood as challenging to them.

A good wording for this is "open-washing" as described in this paper: https://dl.acm.org/doi/fullHtml/10.1145/3630106.3659005

ssahoo2• 2 years ago

Just call it not open-source? May be. Freemium?

maxdo3• 2 years ago

I'm really unsure if it's a good idea given the current geopolitics.

Open-Source Code in the past was fantastic because the West had a monopoly on CPUs and computers. Sharing and contributing was amazing while ensured that tyrants couldn't use this tech to harm people simply because they don't have a hardware to run.

But now, things are different. China is advancing in chip technology, and Russia is using open-source AI to harm people on the scale today, with auto-targeting drones being just the start. Red sea conflict etc.

And somehow, Zuckerberg keeps finding ways to mess up people's lives, despite having the best intentions.

Right now you can build a semi-autonomous drone with AI to kill people for ~$500-700. The western world will still use safe and secure commercial models. While new axis of evil will use models based on Meta or any other open source to do whatever harm they can imagine with not a hint of control.

This particular model. Fine-tune it to develop a nuclear bomb using all possible research that level of government can get on the scale. Killing drone swarms etc. Once the knowledge got public these models can be a base model to get expert-level knowledge to anyone who wants it, uncensored. Especially if you are government that wants to destroy a peaceful order for whatever reason.

rmbyrro4• 2 years ago

You think Russia and China wouldn't be able to steal any closed model for a couple million dollars?

that_guy_iain• 2 years ago

Wouldn’t even need to pay to steal it. FAANG have been shown to be hacked by state actors. China has some of the best hackers in the world.

maxdo3• 2 years ago

stealing a model and building the entire AI community is a very very big difference :

Fine tune, update, Run model without very deep domain knowledge, that's what we receive as an outcome.

If you are a software engineer and you steal a model in some close format of Open AI , you will not get lots of benefits even if you understand the format of that model, it's a complex beast by all means.

This is a playbook how anyone can run it.

So yeah, big corp is evil from one side, but oh well, think of North Korea, Russia etc level of evilness and what they can do whit that.

adhamsalama• 2 years ago

So, only the west should be able to use AI to kill people because they're the good guys?

AlexandrB• 2 years ago

This argument reminds me a lot of restrictions on exporting encryption in the 90s.

wavemode2• 2 years ago

You're vastly overestimating the capability of LLMs to create new knowledge not already contained in their training material.

keepswag• 2 years ago

It was not amazing that the West had monopolies because they are the ones using AI and advancing AI tech to harm people. Im not sure what youre getting at here with that comment

https://www.vox.com/future-perfect/24151437/ai-israel-gaza-w...

https://www.972mag.com/mass-assassination-factory-israel-cal...

https://www.theguardian.com/world/2024/apr/03/israel-gaza-ai...

Gravityloss• 2 years ago

Can't it be divided into multiple parts to have a more meaningful discussion? For example the terminology could identify four key areas:

    - Open training data (this is very big)
    - Open training algorithms (does it include infrastructure code?)
    - Open weights (result of previous two)
    - Open runtime algorithm

jll29• 2 years ago

The question is what is "open source" in the case of a matrix of numbers, as opposed to code.

Also, are there any "IP" rights attached at all to a bunch numbers coming out of a formula that someone else calculated for you? (edit: after all, a "model" is just a matrix of numbers coming out of running a training algorithm that is not owned by Meta over training data that is not owned by Meta.)

Meta imposes a notification duty AND a request for another license (no mention of the details of these) for applications of their model with a large number of users. This is against the spirit of open source. (In practical terms it is not a show stopper since you can easily switch models, although they all have subtlely different behaviours and quality levels.)

abss• 2 years ago

Interesting, but we have to consider this information with skepticism since it comes from Meta. Additionally, merely open-sourcing models is insufficient; the training data must also be accessible to verify the outcomes. Furthermore, tools and applications must be freely deployable and capable of storing and sharing data under our personal control. Self-promotion: We have initiated experiments for an AI-based operating system, check AssistOS.org. We recently received a European research grant to support the improvement of AssistOS components. Contact us if you find our work interesting, wish to contribute, conduct research with us, or want to build an application for AssistOS.

Purplehermann• 2 years ago

Well that's it then, we're gonna die

bzmrgonz• 2 years ago

I see it as a new race to build the personal computer (PC) all over again. I hope we can apply the lessons learned and can jump into open source to speed up development and democratize ai for all. We know how Microsoft played dirty in the early days of the PC revolution.

arisAlexis• 2 years ago

Zuckerberg and LeCunn put humans at great risk

notfed• 2 years ago

Great, maybe you (Meta) should actually release the source for your AI then?

Open source AI is the path forward

Related Stories