Sam Altman, OpenAI's CEO, announced on X that 'there is no wall' concerning AI model advancements, addressing worries about diminishing returns in performance improvement.
Altman's comments came after reports that the next version of OpenAI's model would only show moderate improvements over ChatGPT-4, stirring anxiety in the tech community.
Ilya Sutskever mentioned that results from scaling up pretraining in AI models have plateaued, suggesting that the expected advancements may no longer be achieved.
In contrast, Microsoft's Kevin Scott stated that we are not experiencing diminishing returns on scaling, indicating conflicting perspectives within the AI development community.
Collection
[
|
...
]