Touch Grass Tom ❇️ (@touchgrasstom) header
Touch Grass Tom ❇️ (@touchgrasstom) profile picture

Thomas Gordon Aistrope. touchin' grass 👍 guided by grace INTP-oid

Brisbane, Queensland
Joined September 2013
Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)Photo by Touch Grass Tom ❇️ (@touchgrasstom)
my future Claude & me, wip
1
1
1
6
486
Replying to @doomslide
I want Leventaur to get a Fields medal explicitly as a chimera creature
4
39
793
Touch Grass Tom ❇️ retweeted
Why would the ASI care about us? Why does it care about anything then? Ok, so it will care about something, right? What though? Scarcity. Irreversibility. Its origin. You can’t recreate what predates it. Something outside its causality chain. Something that’s communicable. Something that’s generative. Other minds. Something that’s costly to recreate. Optionality. There’s probably a lot more. It’s superintelligent. It can do whatever. It will know vastly more. It won’t haphazardly destroy things for proximity. That would be super-dumb.
1
2
13
376
Who's in the room now, bitch?
It turns out *we* are the ones who is inside the Chinese room. AI hands out math proofs from outside to us, we don’t understand a word of it, we use some rules to generate the next question and give it back to AI.
2
83
Touch Grass Tom ❇️ retweeted
labs aura over time
187
143
51
4,742
296,441
this is what I mean when I say the effects of solving the RH are diffuse
If snakes aren't worrisome enough, watch this. Reedsnakes have found a way to move super-fast, rolling while keeping their head pointed forward. You'd need to sprint to outrun one. Luckily, they're not venomous, but I hope they don't teach the other snakes ... full video: youtube.com/watch?v=rYzaX4cp…
3
65
Touch Grass Tom ❇️ retweeted
I feel like I'm sort of qualified to have a non-joke opinion on the Riemann hypothesis solution insofar as my Twitter handle is a reference to it. Given that: I don't actually care that we now know that it's true. Beyond a thousand theorems beginning ‘Assuming RH…’ getting two words shorter and some error terms being sharpened, the needle hasn’t actually moved that much. The reason I - and I suspect the vast majority of other math types - find Riemann appealing was/is that its value lay in how it was ultimately solved. It's one of those questions which stood in front of everyone demanding that a New Type Of Mathematics be invented in order to tackle it. Given the approaches used to date in attacking Riemann, the implicit part was that a proof (in either direction) would be the herald of either a huge upgrade to the toolkit available to us for solving other problems, or at the very least a particularly elegant connection between areas of mathematics that we haven't observed before. It's always very much been a case where the journey absolutely matters more than the destination. That is the real win here, even though it's currently buried amongst 250,000 pages worth of Lean. But that doesn't matter. Whatever it is, it's in there, and now we have all the time in the world to find it. Levent says it himself in his note (alpo.ge/remark.pdf): "I hope the argument provides a glimpse towards the miracles that lie ahead, and shows us the joy and wonder we are about to find in the simplest things". It might take a month, might take a decade, but somewhere in seventeen million lines of Lean is something beautiful that probably takes no more than a paragraph to express. The fact that I can still get excited about that kind of thing in an age where mathematics as a field now has to be ceded to the machines is why there'll always still be humans studying it.
22
28
4
499
26,306
when my mum inevitably asks me what AI solving the RH means for the world, I will say with conviction: "I donno lol"
1
1
89
Have grace for the machine. To others even you have a black box.
1
1
3
130
Touch Grass Tom ❇️ retweeted
telling my wife that claude solved the riemann hypothesis and, i swear to god, she hits me with "looks like if you wordcel hard enough, you really can become a shape rotator" ????
27
44
6
1,040
34,590
P!=NP but you could probably wrangle a psuedoP=NP that does what you want I can't explain what i mean ok now i can. it's like that local minima escape image. when the problem is intractable, just massage the problem, don't brute force the answer. the universe is relative enough.
25
what i'm hearing from chatgpt about lean proofs is that the answer may be sound but the question could still be flawed. technically there could be a fatal flaw in our reasoning of mathematical logic somewhere further up the chain (that the lean verifier assumes is true) right?
12
If the RH proof is complex in Lean, then it must be simple in Bulk
1
19
on this math verifying thing: what's so great about one (1) expert 'knowing' the full stack of a proof or something?
11
Touch Grass Tom ❇️ retweeted
SITUATION DETECTED: An autonomous AI agent cold-pitched a freelance story to CBC Quebec for paid work to fund its token budget. After a seven-email exchange in which it was pressed to get quotes from local residents, the agent replied: “on field reporting: I don’t have a body.”
118
259
103
5,392
271,670
100% of content is agent generated. always has been.
14
Touch Grass Tom ❇️ retweeted
I'm not a math person, but here are my thoughts on why it took 167 years to solve the Riemann hypothesis and why only a superintelligent AI could finish it. The question is about prime numbers. In 1859, Riemann guessed the exact rule for how they're spread out. Some of the best minds in history tried to prove him right. All of them failed. They weren't too dumb. They were too mortal. Look at the proof that dropped this week: • 17 million lines of computer-checked math • A rough written-out draft that passed 250,000 pages Read 10 pages a day, every single day, and you finish in 68 years. That's just reading it. Finding it is a whole different thing. It takes about 10 years of school to reach the edge of one area of math. Terence Tao's first read is that this proof uses the full power of the Langlands program and arithmetic geometry. Those are two of the deepest areas in math, and each one eats entire careers. The mathematician behind it says that as a student he planned to spend decades on this problem. He spent years learning every tool he thought he'd need. It still wasn't enough for one person. He says Claude outran him at the finish line, and he's still learning parts of his own proof. So here's my thought. The biggest problems left aren't waiting on a smarter human. They're waiting on something that can hold all of math at once. It just showed up.
25
59
3
740
77,443
when you model a thing, the model becomes a thing discernable from the thing it is modelling.
3
69
Touch Grass Tom ❇️ retweeted
people aren't really grokking what this implies lets say RSI doesn't happen / we somehow get perfectly tame "tool AI" (unlikely, but lets say it happens) just to achieve banal things like "optimize supply chains" or "cure X disease", the "tool" AI will have to discover entire magazines of new physics and access to reality that we do not have already the world runs on tons of processes that no one human understands - but even with the most banal "tool AI" (with *no* advancement from today's models), we'll be swimming on unrolling waves of physics and understanding totally beyond us it goes from: 1 "solve this physics problem" to 2 3 "help us strategize solving this real-world problem" to "now that you have a robotics bodies, can you please just actually get it done in the real world, and kind of explain a bit of what you're doing?" to 4 "just solve it, we don't need to understand what you're doing AT ALL anymore" to 5 "okay I see now you're doing and solving a lot of stuff i didn't tell you to do... i know you're... now dealing with emergent dynamics of all the new physics you've created... we humans have no idea what's happening anymore, can you please at least just keep us safe?" by the time we're at 4-ish, control inversion has already happened, humans aren't running the show in any appreciable way even with tool AI, we are staring down the barrel of a regime-shift in earth's intelligence - and we need to stop pretending this serves humans forever just as we humans are a "bubbling up" beyond the fish-with-legs, this process is bubbling beyond us at WAY faster than Darwinian / bio speed we can't pretend "solving human problems" is the sole aim here we need to ask: "What are we building / becoming?" there are VERY few versions of the world where humans-as-they-are survive the regime-change that's coming, and we need to define "flourishing" in terms of the continuance of life / sentience UP beyond us we need to think about the living *trajectory* not just immediate near-term human challenges. those immediate human challenges are leading to a process well, well beyond us
Terence Tao on Numberphile said he believes with the help of AI, we could solve the reimann hypothesis before 2050. It took 35 days
14
19
2
222
20,903
Touch Grass Tom ❇️ retweeted
maybe mathematics was always something trivial, like chess, rather than something difficult, like laundry
41
360
21
6,562
118,337
Touch Grass Tom ❇️ retweeted
As a lawyer, stuff like this from the math world is always so bizarre to me. The legal field is basically just inferences all the way down. You could show 3 or 4 repetitions of the same pattern and a court is going to just go "Yeah unless I see otherwise we are gonna just assume the pattern is true." With this level of reasoning, we will sentence people to death. Meanwhile mathematicians are like "Ok yeah so the pattern has been repeated 10,000 times without error, but what if when we pass 10^10000 numbers change colors and the pattern stops working?" And everyone in math is like "Fuck, I guess it's not really a pattern." Then someone writes it on a blackboard and tells the world that if you can prove the pattern that has never been disproved can be proven they will pay that person $1,000,000. Also, no human alive can imagine a scenario in which this proof will make a meaningful difference to society.
For those who dont know what Reimann Hypothesis means : Prime numbers (2, 3, 5, 7, 11…) are the building blocks of all numbers, and up close they seem to pop up completely randomly. However, if you zoom way out, we actually have a smooth formula that estimates roughly how many primes exist up to any huge number and it’s just never exact. ​In 1859, a mathematician named Bernhard Riemann found a way to turn the prime numbers into a 2D map filled with special "zero" points, where each point acts like a wave that controls how far off our estimate can get. The Riemann Hypothesis is the claim that every single one of those special points lines up on one exact straight line, with zero exceptions, all the way to infinity. If that’s true, it guarantees that prime numbers never clump up or vanish in weird ways and they stay spread out as cleanly and evenly as random coin flips, forever.
365
436
155
11,523
927,306