OpenAI’s new “o1” model looks very cool and has a different approach than the company’s other model offerings:

We trained these models to spend more time thinking through problems before they respond, much like a person would. Through training, they learn to refine their thinking process, try different strategies, and recognize their mistakes. 

In our tests, the next model update performs similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology. We also found that it excels in math and coding.

Fascinating stuff. o1 is trained on how to solve problems, not just with the world knowledge base of traditional LLMs.

Ben Thompson has a high-level explanation for how the model works on Stratechery:

In summary, there are two important things happening: first, o1 is explicitly trained on how to solve problems, and second, o1 is designed to generate multiple problem-solving streams at inference time, choose the best one, and iterate through each step in the process when it realizes it made a mistake. That’s why it got the crossword puzzle right — it just took a really long time.