ex-programmer, failed Phil/osopher. resident painter ⚡️🤖, card carrying member of the linux militia. been laying groundwork for a punchline that never landed
�̶̡̻͙̜̯͉̰̦̗͖͖̖̊̓͊̀̈̊͊̅͆̍̋̄̽̄͝͝
Joined March 2012
- Tweets11.9K
- Following1.2K
- Followers24.2K
- Likes8.5K
Pinned Tweet
just published "Decentralized crypto needs you: to be a geographical decentralization maxi"
collective.flashbots.net/t/d…
this is true for all formal verification in general, cryptography, game theory, any attempt to capture reality via modeling or mathematics
there are always gaps, and those are more important than the correspondences. for example, in eprint.iacr.org/2023/044 we focus on how proofs of knowledge can be created without knowledge, and what that means
this is true for any formalization, for any math, I have not looked to see if this one is more or less faithful to reality than average, but yes, defining the property and *what you are trying to prove* to correspond to your human intuition is the hardest part of all of this, not actually proving things where the machine outputs 1
BREAKING: OpenAI’s solution to Navier–Stokes does not match its Lean verification.
The most important article to read today is not one of OpenAI’s 700 AI-generated math papers.
It is this other paper, making a deep and worrying point:
A Lean-verified proof does not automatically validate the proof written in natural language, nor does it mean that the formal statement captures the intended theorem.
During translation, an AI can change an assumption, weaken a statement, or replace the argument entirely.
It can hallucinate another theorem.
Lean correctly verifies the result.
But the proved result may no longer be what the paper claims.
This is a general problem. Things get spicy when the authors examine OpenAI’s proposed Navier–Stokes solution.
They identify at least two mismatches between the written intermediate results and their Lean counterparts:
One estimate claims that four additional input derivatives suffice. The Lean version requires five: a weaker result.
A pressure-flux estimate is obtained through a different bound, and proved through a different argument.
It is not clear whether these mismatches invalidate the entire proof.
But they raise an important issue.
OpenAI is flooding us with claimed revolutionary breakthroughs. Yet nobody knows whether the proofs are correct or whether they prove what they claim to be proving.
Epistemia at scale.
*
Paper in the first reply
.... and plzplzplz do not forget to go deep into mev or you will be sorry 👇
xitter.twitterapi.xyz/sarahalle_/status/2077…
there is no excuse, this is your best chance all year! there will be cake yummy delicious cake!
NYS & California anti manufacturing (CNC, 3D printer, lathe, etc) legislation is the #1 defining legal/technical freedom issue in this political moment, IMO. even beyond AI control
tomshardware.com/3d-printing…
in Ukraine, open source drone designs and home manufacturing has played a critical part in the war effort. grassroots knowledge is currently turning the tides against a literal superpower that is willing to burn money and conscripted lives, in one of the most inspiring triumphs of human perseverance of the 2000s
I am no manufacturer of firearms, but the right to make what you want and make your own choices about whether what you build is lawful rather than having it technically imposed on you, is a fundamental human right
and the right to run open source software that bypasses local protections of your choice is also your right. this is already enshrined as speech, and there are decades of ideation on it
without these inalienable technical rights, there is no society. they are core parts of numerous, disparate protections provided by our Bill of Rights
this is before you even consider that policing home manufacturing requires mass surveillance and criminalization of open source software (in which protections can be trivially removed) and even open hardware (which can operate all of this machinery on a plug and play basis)
if you cannot make your own stuff, you are at the mercy of those that can. more reliant on capitalism. weaker in a struggle. less decentralized. it is an enormous L to society, and a stain upon a state I am otherwise proud of that this managed to pass any kind of political process, mostly via lobbying and other forms of payola
you have to ask yourself, why do these laws target equipment like 3d printers, which are not substitutes for gun mills and cannot produce functional pressure bearing parts? one answer is that it is harder to change federal serialization laws. but this seems to me akin to banning hands, because if there are no hands you cannot fire the gun
instead of educating the youth on how to compete on a global scale, reminding New Yorkers how to make things, as a state littered with decaying factories that used to define their country. instead of taking back economic power. we cower in fear of a few un-serialized guns (that are already illegal), in a country where they outnumber people and that will never change
I have been a New York resident all my life. I will never put state mandated software on any of my equipment. my printers, my CNC, my drills, nothing. I believe this is my right, to use the software I develop for the hardware I own. I will keep making things using open hardware and software, same as before the bill
I believe I am a contributing member of society. but, you can come get me if you disagree, @KathyHochul, you have my address.
200% tax on any company, per credit, for any centralized/frontier ai. can scale up further with aggressively high credit/employee ratios. queries exempt for auditing own cybersecurity
spend every penny on public humanistic works and infrastructure. for every lost job, create 2
^-- how to deal with ai job loss progressively without being an n/acc or leveraging the entire economy on a local top (sam altman is the new sbf dont at me)
4
942
noob setup tbh lol
This is where we are right now. And i’m not gonna lie it feels pretty magical 🧚♀️
Qwen3.6 27B running inside of Pi coding agent via Llama.cpp on the MacBook Pro
For non-trivial tasks on the @huggingface codebases, this feels very, very close to hitting the latest Opus in Claude Code, or whatever shiny monopolistic closed source API of the day is.
In full airplane mode.
Most people haven’t realized this yet.
If you have, it means you have a huge headstart to what I call the second revolution of AI.
Powerful local models for efficiency, security, privacy, sovereignty 🔥
2
10
3,365
*consumer* gpus have basically become full computers
some of these things push 1kw. these days handle long standing computational loads by themselves
we used to have a computer, with components, working together. a tight coupling. now we have a few computers, the general one and the parallel one. they kind of communicate
and then apus come in and blur the line even more. we truly live in wild times
1
10
1,431
comparing for my uses gemma 4 26B MoE on a consumer 16GB gaming GPU to qwen 3.5 122B with 5x the context on 128GB strix halo
about 4x speed, 1/2 quality, 1/4 price. not sure if it is useful to me yet
2
4
1,529
insanely good for a consumer product
batteries, batteries everywhere
CATL, the world’s largest battery manufacturer, has unveiled its 3rd-generation Shenxing LFP battery.
• 10% to 35% charge in 1 minute
• 10% to 80% in 3 minutes & 44 seconds
• 10% to 90% in 6.5 minutes.
• Even at temps as low as -30°C (-20°F), CATL said it can recharge from 20% to 98% in as little as 9 minutes
• CATL claims the new LFP battery can retain over 90% of its original health after 1,000 ultra-fast charging cycles
• World-record internal resistance of 0.25 milliohms, ~50% lower than the average
• New self-heating pulse technology for better charging performance in cold weather
• New Cell Shoulder Cooling Technology, which is 20% more efficient
“We always deliver what we promise,” Gao Huan, the company’s CTO, told reporters today at an event in Beijing.
1
12
3,409
ok. i am sorry. holy fuck. I have thought that david kleiman wrote the bitcoin code, since 2013. i read the bitcoin code first thing. it was weird to me
on a whim, I searched him in the epstein files. since back was on the island. and I think he was also involved
my brain is weirdly efficient at seeing edge case human patterns, I am often years early to (validated) conspiracy theories (eg have been angry about epstein since like 2006)
it is pretty obvious to me that d kleiman wrote the code. I read a lot a lot of code. he is the only character even tangentially related to the story that fits
this is before you factor in his death and connection to csw. this all makes you feel bad for/humanize csw too. he was tangentially involved in the greatest invention of the 21st century. probably had kleiman confiding in him.
and he was left with nothing but a story, and no proof. the ego hit would not be survivable for any garden variety narcissist
this is correct. my theory has been the same since 2013
back wrote the paper
finney had the idea (back is not smart enough, finney needed a writer)
cipherpunk list made the connection
david klein wrote the software, a gov windows contractor. that explains c wright fooling gavin
Hi. Professional C/C++ programmer here. The open-source code I can find written by Adam Back and Satoshi Nakamoto don't look remotely similar.
Back's code looks typical of academic Unix programmers who also hack their code to run on Windows.
Satoshi code was written by a professional Windows programmer who also wrote for Unix.
Stylistically, they look nothing alike. There's not enough time between 2005 when I can find the newest Adam Back and January 2009 when Satoshi published Bitcoin/0.1 to account for the change. Both are perfectly competent programmers, but stylistically, they are completely different.
The NYTimes tried to compare their English language in posts/emails. I'm compare their C/C++ language in their open-source code. The NYTimes merely points out they both use C++ as if that's another corroborating detail, when the actual code seems to disqualify Adam Back.
this explains a bunch of unexplained phenomena:
- d klein suspicious death
- satoshi persistent disappearance (h finney death)
- adam back stylistic repeated matched to whitepaper
- craig wright having enough early knowledge to fool gavin socially
- single line c++ file written for windows in an open source cipherpunk project... targetting windows only in a spaghetti style that looks like .NET
when many unexplainable things are easily explained, c'est le truth
