Another very interesting video. Thanks for posting this one, too. It looks to me as though the model got the answer just about perfect, which is really impressive. Can you imagine having been able to ask a computer a question like this two or three years ago? How would we have thought about such a response then? Do you think it would be possible to use this model to do original, groundbreaking research in physics? Perhaps you could point it at some questions that have never been answered satisfactorily (or maybe ask it to think up some good questions that have never been asked!), and see if you can work together with it to produce interesting and novel results. I’ve been experimenting with that a bit myself in my own fields of specialization. I don’t have any results worth sharing yet, but it does seem to have potential as a tool to accelerate the research and discovery process significantly.
Yes. It is a very impressive answer that o1 gives. Jens just wrote me pointing out that in his opinion, o1 is " a very good students" and that it will influence what skills the future scientists need ... similar to what the pocket calculator did. But you are right, what I asked it is not a research level in any account. I have also asked I about some math think for a manuscript within theoretical statistical mechanics and did some impressive integrals correctly. If I had this half a year ago, maybe I could have written that manuscript faster.
Great video. More of this kind of physics testing please. Thx!
a video on a research level question next?
Another very interesting video. Thanks for posting this one, too. It looks to me as though the model got the answer just about perfect, which is really impressive. Can you imagine having been able to ask a computer a question like this two or three years ago? How would we have thought about such a response then?
Do you think it would be possible to use this model to do original, groundbreaking research in physics? Perhaps you could point it at some questions that have never been answered satisfactorily (or maybe ask it to think up some good questions that have never been asked!), and see if you can work together with it to produce interesting and novel results.
I’ve been experimenting with that a bit myself in my own fields of specialization. I don’t have any results worth sharing yet, but it does seem to have potential as a tool to accelerate the research and discovery process significantly.
Yes. It is a very impressive answer that o1 gives. Jens just wrote me pointing out that in his opinion, o1 is " a very good students" and that it will influence what skills the future scientists need ... similar to what the pocket calculator did.
But you are right, what I asked it is not a research level in any account. I have also asked I about some math think for a manuscript within theoretical statistical mechanics and did some impressive integrals correctly. If I had this half a year ago, maybe I could have written that manuscript faster.