dooraven

CTO / new CEO definitely approved these changes, founders don't have 50%+ equity at this stage of the business.

The problem is that if there is no cost to a dispute then people will just keep doing again and again.

What's to stop Russia launching another invasion after 2-3 years if they settle it all online?

yeah most likely they have like 6 billion left in the bank accounts which they'll redirect to the new AI lab

Yeah maybe in North India.

South India it's pretty impossible, since it went through the self respect movements and basically got rid of surnames. Most south indians have surnames based on their fathers names.

But it is definitely common to use surnames as proxies in North Indian communities (unfortunately)

that seems like an awful deal for OpenAI. Especially since there are so many AI Labs popping up, how exactly are they going to make enough profit to pay back 10 billion?

why there is still no rich text element is beyond me tbh

there are so many RTEs, just copy one and implement it. Go through the working group RFCs to get consensus etc

I just watched videos today of an ongoing ethnic cleansing of Armenians in Karbakh that is happening today. Although from the mouths of a lot of western governments this is not an ethnic cleansing but a lot of Armenians spontaneously deciding to leave after a military invasion.

Yeah right now the Azeri government hasn't really done anything yet. Azeris and Armenians absolutely hate each other and Armenians there are rightly fearful of their lives but you really can't intervene on something that hasn't happened yet especially since Azerbaijan agreed to w/e US and EU asked with regards to minority protections etc (though the history of Azerbajian's government makes this questionable).

The Ukraine was fighting to regain its territory prior to Russia's intervention, and Russia was effectively accusing the Ukraine of performing an ethnic cleansing of Luhansk/Dotensk. Whether accurate or not, this was their justification for intervention too.

Yep, and the US barely did anything there. The US provided weak support to Ukraine there but the general attitude was "Russia looking after Russian speaker interests" - the Obama admin in 2014/2015 (and either did Bush / Obama do anything in Georgia / South Ossetia) didn't do much to stop it or help ukraine. Germany / France / Europe were still trading with Russia etc. It's only when Putin decided to invade Ukraine proper did the West react because there really isn't a good justification for invading Kiev

What I'm saying is it seems depending on a country's interests, what is an ethnic cleansing and what is not seems in eye of the beholder.

Generally agree but Serbia refused to compromise on anything (eg they banned the Albanian language in schools). NATO only decided to intervene when the Račak massacre happened and I don't think there is any documented proof of that happening in Ukraine / Karabah right now.

Basically until proper massacres are documented, intervention rarely happens.

Difference is that Serbia was actively committing ethnic cleansing.

That being said I do think Northern Serbia should be allowed to join Serbia and Srpska be allowed to join Serbia too.

The land swap deal was a good idea, shame the EU vetoed it.

that's really cool, your definition of a great entrepreneur is a bit outside the common parlance but yeah I'd definitely agree that your father in law is a successful entrepreneur

Sustainable profitable businesses that make a healthy profit and give back to their communities are the lifeblood of economies

please name some then

are you confusing inventors and entrepreneurs? Cause no way you'd consider Tesla a great entrepreneur, he died penniless and JCR Licklider same thing - great inventor and innovator yeah. Great Entrepreneur no.

Of the 3 you listed only Jobs qualifies.

Also this is just lol

The great entrepreneurs of our world are hardly motivated by money

Yeah come on like Bezos, Musk, Zuckerberg, Gates, Hoffman etc aren't motivated by money lol.

For a forum created by an accelerator, it is weirdly anti-success.

pretty hard to show metrics to investors if you don't capture any of them

I managed to get this which is pretty female looking imo but yeah there is a bit issue of an issue when there isn't many femalish options: https://i.imgur.com/CWca05u.png

Yeah you generally only have 2 or 3. Wouldn’t recommend having more than that

Advisors asking for equity and compensation is standard. Give them one or the other but not both.

Use the FAST agreement to guide you https://fi.co/fast. Do not deviate from this agreement, no matter how good the advisors will be.

They fundamentally are not building your business.

I think this is broadly accurate for consumer apps.

However, a B2B app ,which is where most product demo tutorials are used, isn't like this. Good product tutorials are great for salespeople to walk the customer through the app and sell properly to prospective clients.

When the contract sizes are in the 10,000s or 100,000s and the product is pretty complex then this is a good solution vs having your sales and support staff having to firefight issues.

Big 5 is accurate but hard to explain. MBTI is fairly easy to explain and honestly it's not like astrology where it is random, an INTP does act very differently than an ENFJ

You're looking at it from Musk taking over twitter, which I agree with, chasing away users is easy.

The conversation was about X being built from the ground up and attracting new users which is much harder.

BlueSky only started gaining traction after Musk started pissing people off, doubt prag would have done the same.

Can Language Models Teach Weaker Agents? Teacher Explanations Improve Students via Theory of Mind

https://arxiv.org/abs/2306.09299

TokenFlow: Consistent Diffusion Features for Consistent Video Editing

https://huggingface.co/papers/2307.10373

Need to see code for second one.

Yeah me too but to be fair technical users (we're on HN) are probably not the target end users for the product that integrates this.

I'm curious about the composition of your tech bubble - is it programmers / technical folk mostly? cause yeah lots have moved.

The non-technical tech sphere hasn't really moved though, but maybe my circles are quite different to yours.

If they can replicate WeChat it might be a good outcome, since WeChat is essentially the everything app for most China

Problem is that the CCP is basically intimately tied to it so WeChat has massive advantages there.

It still has a boatload of users and is still the cultural town square of the tech world. To build X from scratch you'd have to convince people like Marc Andreessen, Paul Graham, a bunch of other VCs etc to move.

Changing behaviour is very hard. You'd have to get people to install new apps, get use to new workflows etc and humans are creatures of habit so this is very hard.

yeah it's especially annoying on which CUDA version you have to guess to make it work. Sometimes projects don't even produce conda environments properly.

Based on my understanding after skimming through the paper (and assistance with Claude), the AI did not directly generate any full video content for a South Park episode. It seems like this was how the AI was used:

- Custom diffusion models were trained on South Park character and background image datasets. These models could then generate new South Park-style characters and backgrounds.

- GPT-4 was used to generate dialogue for scenes, based on prompts about the overall episode premise and plot points.

- An "AI camera system" was mentioned for scene setup, but details were not provided on how much of the camera work it handled. Voice cloning was used to generate audio clips of the dialogue.

Note: this is just a skim of the paper, entirely possible I and Claude may have have missed something.

ok now I am confused, as Meta themselves say Palm-2 is better than Llama 2?

Llama 2 70B results are on par or better than PaLM (540B) (Chowdhery et al., 2022) on almost all benchmarks. There is still a large gap in performance between Llama 2 70B and GPT-4 and PaLM-2-L.

https://scontent.fsyd7-1.fna.fbcdn.net/v/t39.2365-6/10000000...

If Google's publically available model is better Llama 2 already then why is it so inconceivable that they'd have private models that are better than their public ones which are better than LLama already.

Palm-2 isn't better than GPT-4 but the convo was about better than Llama models no?

Bard release, it would have made more sense for them to have a more limited release of a better model for PR reasons than what actually happened.

Yes I would agree with you if Google wasn't set on to full on panic mode by their investors about releasing something vs Open AI due to Chat GPT's buzz.

Bard was just a "hey we can do this too" thing, it was released half assed, had next to no marketing or hype.

Vertex AI is their real proper offering, and I want to see how PaLM 2 does in comparison.

We still don't have access to Imagen last I checked, it's still in restricted access. We don't have access to SoundStorm or MusicLM

https://imagen.research.google/

https://google-research.github.io/seanet/soundstorm/examples...

https://google-research.github.io/seanet/musiclm/examples/

Why would it be surprising that they have better models for resarch that they don't want to give out yet?

Google's model is not as capable as llama-derived models, so I think they would actually benefit from this.

Google's publically available model isn't as capable. But they certainly have models that are far better already in house.