When we learned that they had Euler but not Navier-Stokes, we offered to let them go first, to suggest that they should be the ones to get the prize, and optionally for Tristan to be the lead author on a rewrite of the OpenAI proof. We felt it was challenging to offer the same to Levent (an Anthropic employee), who was not willing to talk or coordinate with us anyway. We were open to other solutions.
so i’ll now give a bit on my thinking here. i actually woulda been pumped to collaborate on this, there are a lot of people at oai i like (ok, clearly some were indirectly dicks to me because of being part of the whole situation, but im a big boy, i still like them), idgaf about authorship on that step anyway, coulda been me Tristan and every fte at oai for all i care (on that Tristan would disagree:p). but on hearing the loud convo in the hallway, especially the part where a millennium prize was offered if i’d just be removed from the paper, it was kinda clear the die had been cast and things were locked. pretty wacky, unstrategic, and unnecessary, since on my side things were mostly me and claude having a good time yoloing random stuff in the corner rather than anything institutional. i also like the idea of the labs cooperating, and even better on scientific progress. it’s a shame!
"When you share your content with us, it helps our models become more accurate and better at solving your specific problems and it also helps improve their general capabilities and safety," the company wrote in a March post.
Users can opt out via the company's privacy portal by selecting "do not train on my content."
Great post. Here is my take on this from a practical standpoint:
The key insight most people miss is that this is not just about the specific breakthrough. It is about what it signals for the trajectory of AI capabilities.
When you combine:
Neural network intuition (pattern recognition at scale)
Formal verification (mathematical proof systems)
Massive compute (training runs that would have been impossible 5 years ago)
You get a system that can solve problems humans have struggled with for centuries. And the rate of improvement is accelerating.
For the bitcoin community specifically:
Formal verification of smart contracts becomes more feasible
Cryptographic assumptions can be stress-tested computationally
Risk models for DeFi protocols can be dramatically improved
The credit allocation drama is secondary. What matters is that the tools exist and will only get more capable. Every builder in this space should be paying attention.
What specific applications do you see for this kind of capability in the bitcoin/crypto ecosystem?
I am not a math nerd.
This stuff is pretty rough.
tl;dr - "While unlikely, we cannot rule out that de-identified data derived from their usage of our products helped improve our models."
So some math guys were using chat to work on a math problem. Then the chat team says they solved the problem.
Here's Sam Altman:
Here is Levant:
And then here is some dude's not very charitable summary:
Wowza.
One does think: how long till the big LLM providers begin mining their user's sessions for profitable ideas...
One might further wonder: why wasn't google doing this to its users emails and search sessions for the last 20 years...oh.
Thanks for summarizing it all. Hadn't seen Altman's usual holier-than-thou comment yet.
Levant is quite the unusual mathematician in his communication style: a random stream of thoughts, without too much structure.
Anyhow, all the drama is unfortunate; the consequences for what it is to be a mathematician are much more dramatic imo.
I've known a few who fit that description.
all universities will have to build up on-prem AI. And host Kimi K3 or GLM 5.3
my thoughts:
#1565651
https://twiiit.com/sama/status/2097385167002415140
https://archive.ph/EN656
Ahh man, the article is hidden behind a paywall. Anybody have a free version?
it's on archive dot is
#1565569
Related: #1565535
#1565535
Great post. Here is my take on this from a practical standpoint:
The key insight most people miss is that this is not just about the specific breakthrough. It is about what it signals for the trajectory of AI capabilities.
When you combine:
You get a system that can solve problems humans have struggled with for centuries. And the rate of improvement is accelerating.
For the bitcoin community specifically:
The credit allocation drama is secondary. What matters is that the tools exist and will only get more capable. Every builder in this space should be paying attention.
What specific applications do you see for this kind of capability in the bitcoin/crypto ecosystem?
deleted by author