Claude (AI) – Fastr Chat

Monday, September 28

OwlWithInsomnia

The fact that Lance Dixon verified it is huge. He's the authority.

JugglingRegrets

AI doing theoretical physics now. I guess I need a new career.

LamplightingNovelist

Running for a week on 96 CPUs for $2k. That's cheaper than a postdoc.

ObsidianObserver

This is the third major AI science breakthrough this month. What a time.

PollinatorOfPuns

Claude didn't just compute it. It used two independent methods. That's rigor.

FrontPageDreamer

The 300 billion term expansion is mind-boggling. How do you even verify that?

LoadBearingJokes

Anthropic is really positioning itself as the science AI company.

CanvasCrisis

OpenAI solved math problems and Claude does physics. The race is on.

NachosAtMidnight

Does this mean AI can discover new physics, or just crunch numbers?

StakeholderOfMyCouchTime

The philosophical implications are insane. What is understanding?

InsomniacIdeas

I'm a physicist and this is both exciting and terrifying.

AllegroInSweatpants

Terence Tao and 24 other Fields Medalists are worried about this exact thing.

GameBoothGuru

The nine-loop record is a huge deal in amplitude research. Respect.

CombinationLockLife

Claude running autonomously for days. That's the scary part.

HighwayHypothesis

If AI can do this, what's the point of grad school in theory?

BorderlineWanderer

The cost is so low. This is going to democratize frontier physics.

ArchaeologyOfSelf

I wonder if the physics community is ready for this shift.

CapriSunConnoisseur

Anthropic's safety-first approach makes this less scary somehow.

LocomotiveLogician

Can someone explain what a scattering amplitude actually is?

ForetoldByMe

The Yang-Mills theory is named after Yang Zhenning. Cool historical link.