Hugging Face is open-sourcing the improved reasoning ability of OpenAI’s o1 model, that the latter expects to make money off.

https://huggingface.co/spaces/HuggingFaceH4/blogpost-scaling-test-time-compute?

41 points · 3 comments · view on lemmy.world

3 Comments

BaylorSwift3@futurology.today · 17 pts · 1y (2 replies)

Just as the investor funded AI companies think they have the shallowest of moats, and a chance to build revenue, no - once again Open Source is snapping at their heels. Someone is going to blink first in this race, and I suspect it will be the investor funded AI companies.

Espiritdescali@futurology.today · 4 pts · 1y (1 reply)

Do these open source models have access to the same volume of training data that the commercial models have?

BaylorSwift3@futurology.today · 2 pts · 1y

Paradoxically their approach is to use less training data. HF are saying they have reverse engineered some of the capabilities of OpenAI’s o1 model, by using an approach called 'Test-time compute scaling' which OpenAI have acknowledged using, but not disclosed exactly how.

https://the-decoder.com/study-shows-test-time-compute-scaling-is-a-path-to-better-ai-systems/