zlacker

This will be an interesting test to see how fast you can bootstrap GPT-4 level performance with unlimited funds and talent that already has deep knowledge of the internals. With the initial adoption of ChatGPT alongside Copilot, OpenAI's data moat of crawled data & RLHF is pretty vast. And that's not leaving the walled garden of OpenAI. You can simulate a lot of this using other off-the-shelf LLMs (see Alpaca) but nothing is a substitute for real world observed usage.

In a related note, has this meaningfully broken through to the mainstream yet? If a ChatGPT competitor comes out tomorrow that is just as good - but under a different brand - how many people will switch because it's Altman backed? I'll be curious to find out.

>>icyfox+(OP)
I would switch, but not because of Altman backing or not. I would switch if their strategy were to be to progress at pace. I’m not big on AI safety as it is parroted these days, I just want more AI, faster.

>>icyfox+(OP)
Many of OpenAIs most talented people left to start Anthropic. They have billions in funding and have not yet got particularly close to GPT-4.

I think that illustrates it will be a be a big uphill battle for any new entrant no matter how well funded or resourced.

>>tempus+X4
The parent post is literally true yet keeps getting downvoted — what a mess HN has become, too

>>tempus+X4
noob ai here, but is it gonna be challenging because of something intrinsic to gpt4? or about collecting equivalent amount of data to train a comparable model. Because I see Facebook releasing their models down to the weights

>>tempus+X4
And the new CEO was a consultant to Anthropic, apparently. I'm only grateful I don't have to make sense of this drama.

>>tempus+X4
Speak for yourself, I cancelled my GPT4 subscription because I prefer using Claude 2.

>>tempus+X4
Anthropic was formed for nearly the same reason Sam was fired. To slow the things down. OpenAI takes MS funding and Anthropic is formed. OpenAI pace goes a little above the comfort level of Ilya and Sam is fired. MS picks up Sam and will try to outpace openAI while openAI will put brakes on itself.

>>icyfox+(OP)
Realistically if they start now they’d need to hit gpt 5 like levels not 4.

Still, given the exodus and resources now available I’d imagine pretty fast

>>tempus+X4
> They have billions in funding and have not yet got particularly close to GPT-4.

Wrong. Claude 2 beats GPT-4 is some benchmarks (e.g. HumanEval Python coding; math; analytical writing.). It's close enough. It doesn't matter who holds the crown this week, Anthropic definitely has ingredients to make GPT-4-class model.

This is like comparing similar cars from BMW and Toyota, finding few specific parameters where BMW has a higher score and saying "You see? Toyota engineering is nowhere close".

This actually shows Sam Altman's true contribution: the free version of ChatGPT is undeniably worse than Bing Chat, and yet ChatGPT is a bigger brand.

(And it might be a deliberate choice to save money for Claude 3 instead instead of making Claude 2 absolutely SotA.)

>>leoh+l6
-2 points and a truly shifting +/- situation

@dang — any plans to do anything here

I mean not like you have to but yeah I can think of some stuff that could make this better probably (or at minimum experiments that could be run)

Also not on this post but in general I mean

>>leoh+l6
-2 points and a truly shifting +/- situation

@dang — any plans to do anything here

I mean not like you have to but yeah I can think of some stuff that could make this better probably

I mean not on this post in particular but as an HN issue if we agree it’s kind of degrading the experience and there are indeed likely fixes