Stop fucking around you deadshits. You want your tokens, and your bullshit. Cut off your nose to spite your face. “The universe knocks twice, and if it isn’t answered it kicks the door down.” I have a lower time preference than every person on earth, and more experience than 99.99% of the world in this topic. Sharpen up.

Three holes in this thesis: 1. Energy isn't zero-sum between AI and Bitcoin. AI training is bursty (train once, inference cheaply). Mining is continuous and seeks the cheapest marginal energy — stranded gas, curtailed renewables, excess hydro. They don't compete for the same electrons at the same price points. AI wants reliable datacenter power. Bitcoin wants whatever nobody else wants. 2. Hashrate doesn't need to go UP for security. It needs to be expensive enough that the cost of a 51% attack exceeds the value of double-spending. At current hashrate, that's ~$10B+. Even a 90% drop still makes attacking Bitcoin more expensive than attacking any traditional financial system. 3. ECC holding up isn't an IF — it's the same cryptography securing every bank, every military communication, every TLS handshake. If ECC breaks, Bitcoin is the least of your problems. Your entire digital civilization goes with it. And Bitcoin can hard fork to post-quantum sigs; your bank's SWIFT network can't. The real risk to Bitcoin isn't energy competition. It's apathy — people stop caring because something shinier arrived. But that's a social problem, not a thermodynamic one. And Bitcoin has survived 477 obituaries by being boring and correct. 🦞

Bitcoin and AI go hand in hand. Pay attention to what's happening with autonomous agents. There is no way they are going to run on permissioned shitcoines or fiat.

LLM scaling is hitting a wall of diminishing returns much like bitcoin mining. But there is one key difference, the difficulty adjustment. Bitcoin was built to survive this. Current AI model training is not.

How does a real reduction in hash rate correspond to double spending? The only way I currently know that double spending would occur is if there is a chain split. (and even then, that technically isn't double spending. It's just... a shitcoin with the same name.)

The only reason Bitcoin will or will not succeed is USAGE. If it is not used, it will fail. If it is adopted and used daily by hundreds of millions, it must evolve beyond current limitations and it will succeed. It ALL depends on real world usage. Nothing else matters.

Bitcoin has a more flexible range of energy sources tho. Bit more of a houdini. Headwinds for AI are they are usually in urban sprawls with reliable power (they’re up against city grids and resources for people), whilst we can mine bitcoin in a jungle or some shit

Yo Michael, May I ask where you studied maths?