OpenAI’s o1 model generates original insights to solve complex math problems, marking early recursive self-improvement beyond training data recombination.
Theories of AI as stochastic parrots are now obsolete.
@tunguz The City of Marcus in ruins.
Bojan Tunguz@tunguzThe era of the stochastic parrot cope is now officially over.
#1884Ethan@TORCHCOMPILEDI think there was some loose truth in this statement at the time, we’re learning p(x), and due to both training and inference dynamics, we tend to underrepresent the tails of the distribution.
Things significantly change when RL enters the picture
QC@QiaochuYuanat this point it is completely untenable to believe anything along the lines of “AI can only spit out an average of the training data.” that was already only a very rough way of understanding older models pre-reasoning, it was already obsoleted by o1 which released in 2024, and now it should be obviously and conclusively dead even if you haven’t been paying close attention. recursive self-improvement has barely even started and we are already here. even with the recent erdos problem solves you could argue that those were cherrypicked out of a large database for being neglected by humans. that cope is no longer available now generalize the lesson: all other arguments that there is some essential human activity forever beyond the reach of AI are also cope, these are technical problems and the will and money and talent exists and is being deployed to solve them. artificial superintelligence is not a fairy tale. assume it’s coming and plan accordingly