Probably because LeCun is from there. But top AI talent needs to be paid top cash and the taxes there are brutal for high earners especially.

I think it is also a matter on how the Meta stock comp works - and that people hired during the slump in stock price became very expensive once it came back up.

There are some short term ones but I doubt this will continue, especially for the more powerful models.

It doesn't look good for Anthropic, especially considering they are burning billions in investor money.

Looks like they lost the mandate of heaven, if Open AI plays it right it might be their end. Add to that the open source models from China.

I work at a company that has gone all in on Anthropic, and we're just shoveling money at them. I suspect there are a more enterprises than we realize that are doing this.

When I read these comments on Hacker News, I see a lot of people miffed about their personal subscription limits. I think this is a viewpoint that is very consumer focused, and probably within Anthropic they're seeing buckets of money being dumped on them from enterprises. They probably don't really care as much about the individual subscription user, especially power users.

1. HN is so unrepresentative of real life. You have people on their $20/$200 subscriptions complaining about usage limits. They are a tiny fraction of Anthropic's revenue. API billing and enterprise is where the money is.

2. Anthropic and OpenAI's financials are totally different. The former has nearly the same RRR and a fraction of the cash burn. There is a reason Anthropic is hot on secondary and OAI isn't

OpenAI is dealing with exactly the same energetic and financial constraints as Anthropic. That will become apparent soon.

The other thing is most people don't really care about price per token or whatever but how much it will cost to execute (successfully) a task they want.

It doesn't matter if a model is e.g. 30% cheaper to use than another (token-wise) but I need to burn 2x more tokens to get the same acceptable result.

Codex or the Chinese models

I think it is naive to think the government (US or China most probably) will just let some random company control something so powerful and dangerous.

I think it is naive to think that artificial super intelligence will be controlled by anyone.

If it is smarter than all humans combined at everything why would any humans collectively control the ai?

All the ants in your backyard still make no decisions vs you

You'd probably listen to those ants if they put you in a harness and had a little ant-sized remote control that could just, you know, turn you off.

Depending how long they wait to press that button, they might be surprised how little happens when they do.

Isn't the U.S. government at least completely asleep at the wheel or captured by the very same "random" companies? I realize the administration got all pissy with Anthropic but it sounds like the gov and gov contractors are still using their models.

Yeah but they still (at least to public knowledge) do not posses anything that could be called AGI. But as these capabilities increase they'll probably get an offer they can't refuse sooner or later.

It has been useless for long time when compared to Opus or even something like Kimi. The saving grace was that it was dirt cheap but that doesn't matter if it can't do what I want even after many repeated tries and trying to push it to a correct solution.

Maybe something using AI could be implemented - does a screenshot of your game e.g. every second and if it detects anything that would suggest cheating then it informs some central system and sends it a movie of you playing for the final verdict.

Of course this all is based on the assumption that the local AI can do this fast enough with enough precision.

That runs on client side, so can easily be tampered with (assuming you're suggesting that instead of secure boot chain).

Also cheat HUD elements can be simply rendered outside of game window (or even on another device like smartphone).

Well my suggestion is that the local part (which could analyze everything that is happening locally, not only the screen) would be the initial filter and if it detects any hint it takes some data package and sends it to some centralized online system that would provide a final verdict (and ban if needed).

But as you say if it is local then you can essentially run anything on the computer and modify what is ran on it. That basically means it is impossible to make an anti-cheat that is 100% bulletproof aside from something strange like buying a locked-in camera which you need to place behind you as it records everything you do on the PC and then the AI thing happens as I explained.

Maybe for pro play and tournaments that would be acceptable but not for the average player.

Unfortunately a lot depends on the game and software you're trying to use. There are cases where (especially older) stuff on Windows doesn't work but on Steam it works fine.

Also the hope is that when the Linux share of the market grows and more multi-platform engines like Unreal are used then we'll get native versions instead of using Wine/Proton.

On top of that Windows is now basically unusable in many ways so for me at least there is no alternative (MacOS is really bad compared to a well configured Linux desktop, could never get past it treating me as an idiot).