Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Mira Murati leaves OpenAI

twitter.com

Post reply on HN
Print view View original

241–250 of 644 posts

Re: Mira Murati leaves OpenAI

#241

Post by iLoveOncall »

HeatrayEnjoyer wrote: post #221
iLoveOncall wrote: post #159

Earlier quoted context omitted.

People still believe that a company that has only delivered GenAI models is anywhere close to AGI? Success in not around any corner. It's pure insanity to even believe that AGI is possible, let alone close.

What can you confidently say AI will not be able to do in 2029? What task can you declare, without hesitation, will not be possible for automatic hardware to accomplish?

Easy: doing something that humans don't already do and program it to do.

AI is incapable of any innovation. It accelerates human innovation, just like any other piece of software, but that's it. AI makes protein folding more efficient, but it can't ever come up with the concept of protein folding on its own. It's just software.

You simply cannot have general intelligence without self-driven innovation. Not improvement, innovation.

But if we look at much more simple concepts, 2029 is only 5 years (not even) away, so I'm pretty confident that anything that it cannot do right now it won't be able to do in 2029 either.

Re: Mira Murati leaves OpenAI

#242

Post by reducesuffering »

Former OpenAI interim CEO Emmett Shear on this departure:

"You should, as a matter of course, read absolutely nothing into departure announcements. They are fully glommerized as a default, due to the incentive structure of the iterated game, and contain ~zero information beyond the fact of the departure itself."

https://x.com/eshear/status/1839050283953041769

Re: Mira Murati leaves OpenAI

#243

Post by fb03 »

mcast wrote: post #180
aresant wrote: post #13

It is unsuprising that Murati is leaving, she was reported to be one of the principal advocates for pushing Sam out (1) Of course everybody was quick to play nice once OpenAI insiders got the reality check from Satya that he'd just crush them by building an internal competing group, cut funding, and instantly destroy lots of paper millionaires. I'd imagine that Mira and others had 6 - 12 month agreeements in place to…

Trent Reznor and David Fincher need to team up again to make a movie about this.

I'd not complain if William Gibson got into the project as well.

Re: Mira Murati leaves OpenAI

#245

Post by actionfromafar »

petre wrote: post #219
jsheard wrote: post #24

Earlier quoted context omitted.

> It is hard for me to square "This company is a few short years away from building world-changing AGI" Altmans quote was that "it's possible that we will have superintelligence in a few thousand days", which sounds a lot more optimistic on the surface than it actually is. A few thousand days could be interpreted as 10 years or more, and by adding the "possibly" qualifier he didn't even really commit to that predicti…

> it's possible that we will have superintelligence in a few thousand days Sure, a few thousand days and a few trillion $ away. We'll also have full self driving next month. This is just like the fusion is the energy of the future joke: it's 30 years away and it will always be.

Now it’s 20 years away! It took 50 years for it to go from 30 to 20 years away. So maybe, in another 50 years it will be 10 years away?

Re: Mira Murati leaves OpenAI

#246

Post by iLoveOncall »

ttul wrote: post #223
shmatt wrote: post #136

Earlier quoted context omitted.

I feel like this is stating the obvious - but i guess not to many - but a probabilistic syllable generator is not intelligence, it does not understand us, it cannot reason, it can only generate the next syllable It makes us feel understood in the same ways John Edward used to in daytime tv, its all about how language makes us feel true AGI...unfortunately we're not even close

While it's true that language models are fundamentally based on statistical patterns in language, characterizing them as mere "probabilistic syllable generators" significantly understates their capabilities and functional intelligence. These models can engage in multistep logical reasoning, solve complex problems, and generate novel ideas - going far beyond simply predicting the next syllable. They can follow intrica…

> Generative video models like OpenAI's Sora clearly have a world model as they are able to simulate gravity, collisions between objects, and other concepts necessary to render a coherent scene.

I won't expand on the rest, but this is simply nonsensical.

The fact that Sora generates output that matches its training data doesn't show that it has a concept of gravity, collision between object, or anything else. It has a "world model" the same way a photocopier has a "document model".

Re: Mira Murati leaves OpenAI

#247

Post by actionfromafar »

HeatrayEnjoyer wrote: post #215
xur17 wrote: post #35

Earlier quoted context omitted.

Which is arguably a good thing (having AGI spread amongst multiple entities rather than one leader).

How is that good? An arms race increases the pressure to go fast and disregard alignment safety, non proliferation is essential.

I think that train left some time ago.

Re: Mira Murati leaves OpenAI

#248

Post by mewpmewp2 »

deepGem wrote: post #112
aresant wrote: post #30

Earlier quoted context omitted.

I think the much more likely scenario than product roadmap concerns is that Murati (and Ilya for that matter) took their shot to remove Sam, lost, and in an effort to collectively retain billion$ of enterprise value have been playing nice, but were never seriously going to work together again after the failed coup.

Why is it so hard to just accept this and be transparent about motives ? It's fair to say 'we were not aligned with Sam, we tried an ouster, didn't pan out so the best thing for us to do is to leave and let Sam pursue his path", which the entirely company has vouched for. Instead, you get to see grey area after grey area.

Because if you are a high level executive and you are transparent on those things, and if it backfires, it will backfire hard for your future opportunities, since all the companies will view you as a potential liability. So it is always safer and wiser option to not say anything in case of any risk of it backfiring. So you do the polite PR messaging every single time. There's nothing to be gained on the individual level of being transparent, only to be risked.
Post reply on HN