zlacker

Foundational AI companies fracturing is going to be good for the overall market.

Just like Anthropic spinning off from OpenAI brought some very cool research (particularly liked the functional introspection vs node based stuff), Altman going into his own venture to compete will help diversify the overall space.

And while many seem to be thinking this is going to bode poorly for OpenAI, I think perhaps long term it might be a good thing. Ilya is quite impressive and his having greater influence with Altman gone may result in them exploring more interesting avenues that have long term value which might not have been as prioritized with a focus on maximizing short term rollouts.

While the way the board handled it wasn't great optics, this is probably going to be an overall positive shakeup for AI and LLMs in 2024-2025.

>>kromem+(OP)
It probably is a good thing, not because Ilya is going to safeguard humanity’s future, but because it’s better to have a very competitive field rather than one break away winner. No lone genius will invent AGI. It will be a collective effort of dozens of labs, trading talent and methods. If OpenAI is a few years ahead now, they will quickly lose that edge due to this infighting and others will catch up.

>>kromem+(OP)
I wonder how successful Sam's next AI venture will be without Ilya.

>>peanut+Y6
The problems that had been solved remain solved, no need for a genius for those problems. They will also have solutions that the public doesn't know about yet.

>>kromem+(OP)
The default winners in this race, in the medium-term horizon (~18 months), are the existing incumbents: Alphabet, Microsoft, Meta. OpenAI was positioned to be able to compete. I am not sure a diluted OpenAI will be.

>>Palmik+Ab
But I think this episode shows how dependent OpenAI was on Microsoft. They seem to be closer to an extension than a truly independent actor