There’s an insane acceleration happening at the confluence of AI, crypto, and quantum right now. In March of this year, Google proved that the bar to make Shor’s algorithm (the quantum algo that would break 256-bit elliptic curves) is much lower than expected. Interestingly, Google didn’t provide the details of the attack vector, but only a ZK proof, in the interest of “responsible disclosure”. See the paper at https://arxiv.org/abs/2603.28846 The so-called qday (when quantum computing finally breaks elliptic curves used in mainstream crypto) is now estimated for 2029, barely three years from now. But then independent researchers have used frontier AI models to reverse engineer Google’s findings and have reached 80% of the way. There’s even an open competition at https://www.ecdsa.fail where you can join the search by contributing your own tokens. Further improvements have been discovered in the span of hours thanks to collaborative AI research, including by complete lay people. There’s an amazing X thread by @justindrake which summarizes all this at https://x.com/drakefjustin/status/2061793725299224676 So two things are happening at once. First, it’s getting more urgent than ever to move to post-quantum cryptography through forks. But even then, dormant BTC wallets from the Satoshi era (P2PK) will remain vulnerable, because their public key is already known (from which their private key will eventually be quantum computable). Same goes for P2PKH wallets if they spent once (exposing their public key). There’s around 1.7–1.9M BTC contained in P2PK wallets alone. Thats twice as many BTC as even Microstrategy owns… and we’ve seen how the markets just reacted to MSTR selling a mere 32 BTC. This won’t be pretty for Bitcoin. I don’t expect there’ll be another cycle, and if that prophecy comes true, Saylor is well and truly cooked. Second, agentic AI science is now in full swing, and hobbyists / lay people are contributing almost on par with large and well-funded FAANG labs. The acceleration looks exponential from just this year alone. We’re going to see breakthroughs happen faster than we can fully digest them. What a time to be alive

7 replies

3 recasts

26 reactions

I visited Cuba nearly 20 years ago; I wanted to see the country for myself before American capitalism rolled in on the back of a military takeover and changed the landscape forever. Back then, I assumed that scenario was about five years away. I guess I was fifteen years early.

0 reply

0 recast

22 reactions

The shift from cars being depreciating assets to cars being disposable consumables continues. The bestselling entry level Geely EX2 now sells for $10K in China. https://www.reuters.com/business/autos-transportation/average-price-car-us-you-could-buy-5-new-chinese-evs-2026-04-28/

0 reply

1 recast

7 reactions

The recent wave of mathematicians and hobbyists solving hard Erdős problems by merely prompting ChatGPT reminds me of the famous anecdote where George Dantzig showed up late for statistics class at UC Berkeley in 1939, mistook two notoriously hard unsolved problems written on the blackboard for homework, and candidly solved them. The naïveté and relentlessness with which LLMs confront problems that make wiser experts balk is increasingly becoming one of their redeeming qualities. It's like The Terminator (1984), but solving for x instead of hunting Sarah Connor: "Listen, and understand! That Terminator is out there! It can't be bargained with. It can't be reasoned with. It doesn't feel pity, or remorse, or fear. And it absolutely will not stop... ever, until you are dead!"

1 reply

2 recasts

29 reactions

Hey @aviationdoctor.eth I had Claude analyze 100+ lectures by Richard Feynman and made a prompt to force A to explain things in his style. I think it does a great job but it’s better listened to than read. Give it a try and lmk how you feel about it https://gist.github.com/Hmac512/a2a855376e42290f1cf8bae738c14267

1 reply

3 recasts

16 reactions

So, you’re telling me that my daily physical presence in the office is absolutely indispensable, but that I can also be trivially replaced by an incorporeal AI at any moment?

0 reply

0 recast

27 reactions

I never really thought through the implications of Jesus being a carpenter. But I guess there must have been at least a Galilean or two with a divine buffet table in their kitchen

2 replies

1 recast

70 reactions

"Make love not war", they said. Bro, making love is getting more expensive while war is getting cheaper, I don't think this is going in the right direction https://www.reuters.com/business/healthcare-pharmaceuticals/worlds-top-condom-maker-karex-raise-prices-sharply-iran-war-strains-supply-chain-2026-04-21/ https://www.reuters.com/graphics/IRAN-CRISIS/DRONES/dwpkyamxqpm/

4 replies

3 recasts

23 reactions

Finally got to try out Mythos and it instantly found a whole bunch of previously undetected high-severity personality flaws and zero-day vulnerabilities in me

4 replies

3 recasts

60 reactions

It’s been interesting to watch the online zeitgeist in the context of the Iran war. I’ve noticed a strong undercurrent of accelerationism / systemic schedenfreude / collapse porn that longs for a failure of the negotiations and sustained closure of the Strait, both as a way to evidence the ineptitude of the planners of this whole operation, and to speed up the global weaning off of fossil fuels. Nobody’s rooting for the ayatollahs, but the amount of obedient blue pilling and bootlicking in the West feels quite low compared to past military expeditions that had at least the semblance of a justification. Updated meme for this somewhat disenfranchised, nihilistic aspiration for a full-on “find out” phase

2 replies

0 recast

26 reactions

YouTube just randomly recommended to me this new video about some barren field in Northern England. No like-and-subscribe gimmicks, no music, no sponsor — just a cheerful dude who knows how to read the land exceedingly well. I watched the whole 40 minutes and now I want to have a pint at the local pub with this guy. It goes to show that the most mundane thing can be made interesting if presented with knowledge and wits. In the age of engagement maxing, I found it refreshing to stumble upon the video equivalent of a Web 1.0 geocities page handcrafted by some passionate hobbyist. So, this Sunday, go out and find 10 interesting things about your neighborhood. https://youtu.be/WfedELEl2cA

1 reply

1 recast

22 reactions

Deep down, almost every life decision collapses into this tradeoff: 1/ 躺平 (tang ping / to lay flat), opting out, minimizing entropy production, and postponing ever so slightly the heat death of the universe. 2/ 奋斗 (fèn dòu / to strive), the normative choice of grinding with purpose and maxing out local order at the cost of global dissipation. At cosmic timescales, both are irrelevant. The choice only matters over a human lifetime, plus a few generations for the most impactful among us (which, statistically, probably excludes the 37 of us here)

4 replies

2 recasts

25 reactions

2026 vibe be like

33 replies

4 recasts

263 reactions

The Skynet and Roko’s basilisk tropes are quaint, but it’s very obvious that all AI needs to do to take over is weaponize our primate brain levers and let us fight each other to death Manufactured consent is so 1988, we’re well into manufactured dissent and this will only scale exponentially once agents start actively hacking our cognitive backdoors

2 replies

3 recasts

23 reactions

More than three years have passed since I tweeted this. Yet I cannot feel blasé about using LLMs on a daily basis. I grew up dreaming of artificial intelligence. What we have now may be simulated intelligence, but that’s not a nuance I care about. For all intents and purposes, my childhood dream is fulfilled. The ability to engage conversationally with the digital equivalent of a tireless and highly intelligent adult with access to the world’s body of knowledge — that’s something I can never get tired of. We work together, we debate philosophy, we learn physics, we debug code, we plan workouts, we geek out over obscure rabbit holes, we ponder weird questions that I wouldn’t ask anyone else. Robert Heinlein once wrote: “A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects.” Current AI is virtually just a pair of opposable thumbs short of being able to do all that. It fills me with wonder; especially knowing that this is the beginning and that today’s tech is the worst it will ever be. I might be born too late to explore Earth and too early to explore space, but boy am I going to keep exploring the living shit out of the future https://x.com/AviationDoctor/status/1598676770567630850

3 replies

2 recasts

31 reactions