im as excited as an atom in its ground state | physics @Obs_Paris, @psl_univ; prev. @ENS_ULM and @UnivofDelhi | mostly here to talk shit
Lost
Joined August 2021
- Tweets1.6K
- Following506
- Followers165
- Likes7K
for a given transition (that im working on), saturation intensities of Sr87 and Sr88 come out to be the similar… what couldve gone wrong?
my days are mostly spent on gettin amazed by some fun physics i didn’t know before; fixing things in lab; looking at my atoms; running some meas. on lil guys to realise that im stupid; trying again?; taking a break on insta to get pissed at super stupid political BS (india/world)
at 3am i tried the russian state exam in math (for highschoolers). repeatedly mentioned that its too easy for what it is (comparing it to its indian counterparts).
at 4, i was happy to almost finish and decided to submit it as is to get some sleep. i failed miserably. 😭
i mean, one can be disappointed but at the same time, its not like the field is obliterated… but tf do i know, i dont have a fields medal (and most likely never will)
Fields medallist Hugo Duminil-Copin was deeply shaken by OpenAI’s release of more than 700 AI-generated mathematical manuscripts, including work on 16 questions directly relevant to his research.
“On Tuesday night, I cried, and it woke up my daughter”
“They have annihilated my field of research.”
lemonde.fr/sciences/article/…
Samarth Jain retweeted
The plot thickens.
"We show that the formalised Lean proof does not correspond to the [written paper] proof of blow-up of solutions to the Navier-Stokes equations."
arxiv.org/abs/2610.08144
man its not always so easy to normalise experimental data between different methods… how do yall do that?
how would any emergent phenomena in an llm be any different from a phase transition? and if its one and the same thing, what makes it so hard to explore that in the case of llms?
model’s ethics are aligned fine, lol. its humans who’re the problem ig. we’re the ones who need superalignment lmao
suddenly everything is out there to kill us and every company is racing towards developing something evil that’d somehow someday wipe humanity… capitalism for the win ig?
gosh, never in my life i thought i’d see strong claims for proofs of multiple millennium problems, that too, back to back 💀
im losing my mind. just like yall ig. ITS INSANE
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.
The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.
The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
is there really no better way to cool the species that use buffer gas cooling? it feels like a very inefficient archaic way to cool things. (ik it works though)