magic_hamster

This was a fun read, however, I feel that the actual essence is quite thin.

It can be summarized in a short sentence: The cost comes from being always prepared with teams and equipment, and few privately insured riders subsidizes everyone else, getting very high surprise bills - this can and should be a shared community tax which would effectively allow free ambulance rides to almost everybody.

It's an excellently written post though. I suspect people read this author for the experience, more so than actual information. I just wonder if they should just do prose and go wild with their skill instead of coating a few predictable facts with so many layers of color.

If the problem is that there are too many AI slop papers, then requiring real human input and feedback can help reduce these numbers. Let's say you have to defend your paper in person before you submit it, it will cut slop submissions dramatically. The current system is just not built for a world with LLMs producing mountains of content with automation.

Unfortunately people still think BDS is relevant in any way.

Isn't this hilarious:

A patent as well as a design and model registration have been made to protect the technical architecture and design of Openprinter.

So, open as in OpenAI.

the digg/reddit comment tree format is a clear improvement

You are talking about content discovery. But for a sense of community I have never seen or been able to replicate the feeling of belonging and building a shared experience like old crappy forums.

Solving the paper submission is easy. Just hold frontal interview where the submitter defends their paper. They can't create papers every day and still be knowledgeable about them in depth.

We are hurling to a reality where the only noteworthy metric is human to human validation.

The platforms are now flooded with AI slop

What did we do before "platforms"? This wasn't that long ago (20 years give or take). If you want to share, join a club, meet people, do your hobby with a group. Personally I create for the joy of it so I don't really care what's going on on "the platforms".

The jury is still out on whether the advantages and opportunities of AI outweight all the negative sides.

The jury is extremely clear on the benefits of AI, even if it is very annoying and being pushed everywhere. We are not going back to a world without AI, and this is something you just have to accept.

Why does it have to keep that way?

Because you can't put the genie back in the bottle.

Why should artists have to put up with this gross violation of copyright on a massive scale?

I don't think this is necessarily true anymore. Adobe is training their models purely on legally licensed material, for example. Many of the open source models aren't available for commercial work by default. My guess is the copyright issue is not why people hate AI. People hate AI because it's a replacement for humans and for human creativity, which sucks. But it can be legal.

Did anyone ask us if we actually want this?

Who is us? I want AI models. Lots of people want AI models. It's not all nefarious. You can have some fun with AI as a hobby and see what it enables you to create.

As the people we always have the right to say: "this is not ok and we demand change"

Okay, what kind of change do you want?

There are a few issues here that should be addressed. There's the AI hype and deafening echo chambers, but then there's also the actual value you can find in AI when you just try it a little bit quietly.

I totally disagree with the comparison to something like NFT. While AI is being pushed aggressively and it can definitely be annoying, AI is actually useful unlike NFTs.

Much like the author, I also enjoy photography, graphic design, and other creative hobbies. It's entirely my choice how much and where to apply AI.

We have to accept that yes, it's useful, and yes you can definitely produce good deliverables with it, not just slop. Yes, when looking for assets and not the artistic process, many people will use AI and the cost of creative work will plummet. It's not great but that's the way it is.

But it's not like we should just stop creating, especially if it's a hobby. Do it for fun. Or, maybe use AI to try something new.

Either way AI is too useful, it's here to stay and it doesn't need a do-over. It's true, we should accept the world is changing, and no amount of moaning or complaining will make this disappear.

There's a lot of offensive security talent, but this has nothing to do with Palestinians. Israeli intelligence is very advanced and is why Israel has been able to eliminate the leaders of Hezbollah and Iran.

Not everything in Israel is about or related to Palestinians. The Palestinian bias only exists in circles where every thought regarding Israel is immediately evoking a Palestinian connotation. In reality, most Israelis never interact with Palestinians.

To suggest that a sector of Israeli startups exists on the experience of people "suppressing Palestinians" is definitely biased, absurd, and is a slippery slope.

While I share the sentiment, this feels like an extreme, nuclear reaction which might be irreversible. I understand the fatigue, and resentment, but if you are about to be a family, you are going to find that typewriters aren't the acceptable mode of communication nowadays, and that you need some money to raise children.

Even if you are already wealthy and don't actually need to work anymore, going off the grid completely is still the wrong move. There's a lot of ways to spend less time online, improve your privacy and reduce tracking, and still benefit from some of the actual, real advantages of tech.

And the last and maybe most important thing is, we are currently on a roller coaster of disruption and frankly some daunting prospects - but we don't know what's right around the next turn. What the development landscape might be like in a few years, or maybe what kind of new problems will emerge that are not yet clear.

The right move is to take some time off, clear your head and decide if you stopped liking tech altogether, or you just needed a break. If you still like problem solving, limit your AI use, stay effective and skillful, and find ways to enjoy your skill.

I've never met an engineer who actually stopped enjoying problem solving.

Nice writeup. It goes deep into Gnutella, but it's also worth mentioning the slew of sharing programs back then, which was truly like the wild west. Napster, Emule, DC++, Kazaa, to name a few. On many of these networks it was possible to literally browse other people's sharing folders, find really cool stuff, and maybe make some guesses on the what this user was like.

I don't see any value discussing which set of rules and outcomes is better because it's all nonsense.

To be honest, I did have some very in depth conversations with religious friends, and it was enlightening. I am convinced that for them, it has some benefits.

But I know all this stuff is totally made up, so even if you can channel it to a good thing, I personally can never do it. And even though I won't say this to their face, part of me thinks, how can you believe this stuff.

Arguably, all the industries that you mentioned (clothing, food, automotive) have the same symptoms, doing everything possible to increase growth even (and often) at the expense of shipping worse products. At least, this has been my experience with clothing, electronics, appliances, and honestly almost everything. It's very hard today to find good long lasting products. A couple of decades ago you could expect your purchase to last a while, today - hardly.

My point is AI is not going to be built to "benefit humanity" because that's not the incentive in our economy. AI might give us some benefits, but like all tech products currently, it will be designed to benefit corporations and shareholders. It is what it is.

In my way of life, the idea that people follow and care deeply about what some mullah has to say is very foreign. There's a mass of these people though. Their life must be so incredibly different than mine, it's just hard to fathom. I can't even imagine caring about the Pope or what they have to say. In my imagination the Pope is something out of roman times, it's just so weird this still exists today.

Good luck with that. Capitalism doesn't work that way. AI will make money for some companies, but as always, it will be on our expense, not for our benefit. We will get some convenient features, we will grow dependent, and eventually subscriptions will be squeezed as far as we are able to pay, advertising will take over, we will have less choice and worse service.

By then we might not even have computers anymore, or we might have "transparent" computers, i.e. have everything on the cloud and just tell our AI agents what to do.

Sorry Pope Leo, things are not going to suddenly turn into a wonderful utopia, but maybe buy some stocks so you can at least make a buck from what's coming.

Computers still do the exact same thing they did back then, which is to read opcodes and do binary math. That is all. It is people and corporations that found a way to monetize them which is hostile to us users. But I think that with time, this too shall pass and computers will still do what they always did. You can still enjoy designing a small PCB, working with microcontrollers and building something fun.

My view is write the code that matters to you and that you want or need to be proficient with. If you need to defend, explain or discuss code, you are better off writing it yourself.

Education is also figured out. You just need to learn, do and practice for yourself. Telling the agent "to just do it for you" is tempting, but it's not learning. You need to be deliberate when you're trying to actually learn and internalize.

Also, you could spin up your own educational agent with very strict instructions on guiding the user instead of just doing the work. Of course you can always go around it but if you're making an effort to learn, this is a good middle ground.

Let me preface my comment by saying I also still write a lot of code by hand - especially when it's something I know I need to understand in depth, and in some cases defend.

With that said, this caught my eye:

AI gravitates toward single-struct-holds-everything because it satisfies the immediate prompt with minimal ceremony.

This is too general. "AI" is used here as a catch-all, but in fact, it was the specific model under the specific conditions you ran your prompt, including harness, markdowns, PRDs, etc. So it's not fair to say "AI does X!" in this case.

It's also very much up to you. It's very common to have a frontier model plan an architecture before you have another model implement code. If you're just one-shotting an LLM to do everything you get mediocre, more brittle code.

This stuff is still being figured out by a lot of people. But I feel the core of the issue is not using AI well. Scoping, task alignment, validation, are crucial.

The battle is lost. You never had a chance. There's nothing you can do against the constant torrent of AI content that's only getting started. The online communities that we know and love are going to change and there's nothing we can do about it. You can't keep AI out of any platform no matter what the community guidelines say or even if it seems locked down with no bot access.

The only solution is in person meetups, bringing back the 3rd places, joining a club. Maybe it's not such a bad outcome.

Best comment I've read this week.

It will be a beautiful day when I can finally lose all my Adobe accounts and software. Kdenlive is definitely on the right track BUT having a real risk to lose my project after days and weeks of work is not something I am able to afford. I am following this with great interest and waiting for the right time to jump on board.

Good job. llama.cpp is already much better.

These are thoughts of someone who's very good at putting words together, but sadly has little experience with the subject matter.

I’ve thought about this a lot over the last few years, and I think the best response is to stop.

This is exactly where it shows.

LLMs, agents and whatever comes next are not only the future of tech, but they are going to be national resilience drivers for the countries that will be able to support them with power, water and science.

Who is supposed to stop? The US? China? Russia? Everyone? Of course this won't happen. This is an arms race.

But even if it weren't, stopping is the wrong answer. You don't have to outsource your thinking, writing or reading. How you use LLMs is entirely up to you.

There is a way to use LLMs which is beneficial. I treat them as a private tutor available to me for questions. This solved a lot of friction I had with my relationship with LLMs.

More telling is that the author mainly thinks about their relationship with LLMs while in reality the space has moved on to automation with agents. You don't interact with LLMs as much as before, and if you still do, then soon you won't.

Ahents are not really ML. It's harnesses and parsing and memory and metrics. It's software. Should we stop this as well?

Ollama is the worst engine you could use for this. Since you are already running on an Nvidia stack for the dense model, you should serve this with vLLM. With 128GB you could try for the original safetensors even though you might need to be careful with caches and context length.

End of the PC era, there's nothing to tinker with anymore. And certainly no gradient for entrepreneurship for once-skilled labor capital.

This one seems too far fetched. Training models is widespread. There will always be open weight models in some form, and if we assume there will be some advancements in architecture, I bet you could also run them on much leaner devices. Even today you can run models on Raspberry Pis. I don't see a reason this will stop being a thing, there will be plenty of ways to tinker.

However, keep in mind the masses don't care about tinkering and never have. People want a ChatGPT experience, not a pytorch experience. In essence this is true for all tech products, not just AI.

docker-in-docker solution

Goodbye Mac users.

gone are the days of PCIe.

My GPU, NVMe drives and motherboard might disagree.

I see AI as a new, unreliable resource that I can try and tame with good software practices. It's an incredibly fun challenge and there's a lot to learn.