Dan (@kagankans) header
Dan (@kagankans) profile picture

senior fellow at Understanding AI. recent: profile of Hans Moravec in @nymag https://xitter.gq/t.co/sJ7il79RyK

Joined March 2013
Photo by Dan (@kagankans)Photo by Dan (@kagankans)Photo by Dan (@kagankans)Photo by Dan (@kagankans)Photo by Dan (@kagankans)Photo by Dan (@kagankans)Photo by Dan (@kagankans)Photo by Dan (@kagankans)Photo by Dan (@kagankans)Photo by Dan (@kagankans)
Pinned Tweet
The lesson for nonphysicists: under AI, everything turns toward interpretation
Human physics will continue. It just might start to look more like biblical studies or art criticism — fields where the focus is on arguing meaning rather than making discovery. understandingai.org/p/i-spok…
2
227
Yesterday I published something about the effect of AI on theoretical physics, with a focus on quantum gravity understandingai.org/p/i-spok… Today Claude solves two problems in quantum gravity anthropic.com/research/claud…
6
22
5
122
12,481
OpenAI hints at more coming xitter.gq/tszzl/status/212225808…
watch this space
1
3
346
Announcement for the Alignment Institute for Fundamental Science, a group of physicists trying to channel human-AI collaboration. Confusing name! Seems to have nothing to do with "alignment" as usually understood aifscience.org/
140
Human physics will continue. It just might start to look more like biblical studies or art criticism — fields where the focus is on arguing meaning rather than making discovery. understandingai.org/p/i-spok…
1
1
5
407
Great article. We recently had @karch_andreas on the @KnowmadsPodcast speaking about Physics in the age of AI, check it out.
1
2
193
Lovely article. Not just because I am mentioned :)
Yesterday I published something about the effect of AI on theoretical physics, with a focus on quantum gravity understandingai.org/p/i-spok… Today Claude solves two problems in quantum gravity anthropic.com/research/claud…
1
2
12
1,047
AI is driving a rapid increase in theoretical physics papers — submissions to arXiv are up 30% year over year. One physicists told @kagankans that the papers are "poorly written, but I think the science is probably pretty solid." understandingai.org/p/i-spok…
2
2
7
1,929
Some of the most prominent theoretical physicists, including Juan Maldacena, talked to @kagankans for his new piece on how AI is transforming physics. understandingai.org/p/i-spok…
2
1
6
1,675
Dan retweeted
Replying to @synthwavedd
Honestly, if they shit the bed again today I am done. Cant handle it anymore. Will cancel my sub and move the fuck on with life. I really really hope they surprise us all but lately they have been very lame and kind of cringe with optics and releases.
3
1
40
3,083
It is always a joy to go to the opera but also always a reminder of how many deeply crazy people go to the opera
15
6
1
122
4,894
If you smash Vernor and Vinge together you get Verge. Makes you think
94
extremely long abcedarian poem about AI and God and love written by a physicist sounds like a cursed genre . . . and yet this is quite good aronwall.substack.com/p/natu…
72
The issue OpenAI is having with its models is the same issue the world is having with OpenAI (this goes for models and companies beyond OpenAI) understandingai.org/p/openai…
1
161
conspiracy physics reaches toward AI
String theory is what happens when an intelligent system is starved of sensory input and starts hallucinating. Intelligence needs grounding. For this reason, intelligence is practically never the bottleneck to solving impactful problems. The bottleneck is iterated contact with reality.
87
Bumping this since haven’t seen it discussed elsewhere. There was a huge freakout about Astra doing this, on grounds it undermines our ability to understand chain of thought and how well aligned a model is. For example astralcodexten.com/p/the-spe…
apparently the new Gemini is looped three times fundaai.substack.com/p/colum…
4
200
apparently the new Gemini is looped three times fundaai.substack.com/p/colum…
5
23
5,109