The forecast disputes Gary Marcus's fourth-place OpenAI prediction
@GaryMarcus I didn't predict that they are falling to 4th place
more like that Google and xAI could catch up and they are all tied 2nd, if OpenAI can't train a 10T model
Below @scaling01 is predicting OpenAI to fall to 4th place by the end of the year, beyond Anthropic, Google, and SpaceX.
I am not sure about SpaceX pulling ahead of them (but not impossible), definitely think OpenAI is slipping.
Massive fall from grace.
Users dismissed predictions that Google and xAI could tie or surpass OpenAI in AI rankings because they view those firms' models as severely flawed due to horrific training.
No Digg Deeper questions have been answered for this story yet.
in the comments @scaling01 clarifies that what he is expecting is a three way tie for second
Below @scaling01 is predicting OpenAI to fall to 4th place by the end of the year, beyond Anthropic, Google, and SpaceX.
I am not sure about SpaceX pulling ahead of them (but not impossible), definitely think OpenAI is slipping.
Massive fall from grace.
@scaling01 @GaryMarcus Google and xAI's models are such dogshit.
It's not just compute: their training has been horrific.
Having said that, someone should just buy OpenAI and call it a down.
@scaling01 @GaryMarcus He thought your initial was Ant, Google, OAI and SpaceX instead of Ant, OAI, Google and SpaceX
@GaryMarcus @scaling01 Owning the dev tool ≠ having the best model. Between Google and OpenAI based on what exactly? A code editor and a 1T param model? Distribution leverage is not inference capability.