Moon Panda Inu 🇺🇸

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@Noahpinion Ed Witten. I had so much respect for the dude but yet another demo that intelligence and ego and self interest are all separate axis. Just let it go. You will be remembered as one of the greatest mathematical physicist. Your impact may just seize moving forward.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@Frances01896069 Perhaps the intended audience is not mathematicians at all, but engineers and physicists (or another AI) trying to apply the results to something practical and useful.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@trq212 @GustavoValverde This is the way 👆

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@jukan05 If China is allowed to infringe on IP and get away with it, I don’t see any reasons why SPCX can’t

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@OutFrontCNN @katie_polglase @CNN is trash

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@ophello @0x_impai Err… disruption mitigation?

Tokamak being a solved problem is far from reality.

Not saying it can’t be solved but saying it is, is a false statement.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@Kritiisayss Nobody cares about coders unfortunately

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@onejasonknight This AI hate is getting a bit lame. Opus5.5 quasi guided by dev writes astronomically better code than good devs alone. It’s also just going to keep getting better.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@LewisCTech I create scientific software and end users are other scientists but the impact has been absolutely phenomenal. Went from a janky slow poorly written and minimal feature Boltzmann code in a box geometry to a fully unstructured grid, multiphysics GPU parallel code in <2 months.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@RandPaul Well. Allowing funds to carryover will fix most of this problem.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@larry0x Using Sonnet5

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@larry0x On Claude code, with good harnesses, (superpower, karparhy, caveman, grill-me), I can get away with low effort on most things.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@8bit5_0 I think this is exactly the looking at the tree and missing the Forrest moment. The individual code and function is the tree, but why would people not think that collective complex behaviors cannot emerge from bunch of individually simple machinery? Stat mech, people?

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@alexkehr Huh? lol no I want a “book me a trip” sometimes because I don’t have any energy left at the end of a grueling work week to browse around and day dream and do all these stuff select few privileged folks can do.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@v0xium Although chain of thought didn’t just appear out of no where, and it was put into the architecture deliberately, it, becoming useful only after large enough parameter size like an emergent phenomena is hinting at something akin to consciousness. Or at least some collectivism.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@v0xium Yeah the last bit of “given time to check out the code” is definitely no bueno if you guys aren’t even allowed to do that. It takes away the joy in the process, leading to folks leaving, then, when AI finds an edge case it can’t solve in its own, it’ll all end catastrophically.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@rohanpaul_ai EXACTLY. I’m working so much more with AI because I can get 30x more done with it. Pace of progress is going to be bananananana

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@Plinz @RBehiel Nah, mathematicians are generally arrogant pricks who genuinely think they have some unique cognitive skills that make them irreplaceable.

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@Plinz As a physicist/applied mathematician who uses AI every day to solve problems, I approve the use of AI to solve any of the great unsolved physics problems. It might turn out that many were no problems to start with (e.g., dark matter)

[Moon Panda Inu 🇺🇸 @PanDog64356596]

@haider1 Let’s most definitely keep accelerating. Democratize thinking. Stop hoarding progress to the very few elites.