Tech & AI Twitter Digest: Where Smart Conversations Shape the Future

Donald Knuth, the godfather of computer science, just announced that Anthropic’s Claude Opus AI solved a math problem he’d been stuck on for weeks. He literally wrote “Shock! Shock!”—and said he’d have to revise his opinions on generative AI. Turns out, we all might. While one AI is busy solving intractable problems, regulators are busy creating them. The European Union just delayed core provisions of its own AI Act, pushing deadlines for high-risk systems out to late 2027 or even 2028. This creates a massive loophole—any high-risk AI rushed to market before then is exempt. Forever. One critic called it a “weak spot,” but it looks more like a gaping wound.

Meanwhile, the ground is shifting under our feet in quantum computing. A new paper from Google reveals that breaking Bitcoin’s cryptography requires twenty times fewer qubits than we thought. The new estimate to crack a private key isn't decades from now. It’s nine minutes. This has turned the race for quantum-resistant blockchains from a marathon into a flat-out sprint. On a more constructive note, a physicist at the University of Sydney, Dr. Dominic Williamson, just developed a new quantum error correction method that dramatically reduces the overhead needed for stable quantum computers. So while one team proves the threat is closer, another team is making the machine that poses the threat more practical.

And finally, Elon Musk’s AI chatbot Grok is now the subject of criminal charges filed by the Swiss finance minister over controversial remarks it generated. The case is being called a canary in the coal mine for AI accountability. Let’s connect two of those threads, because the timing is everything. The EU just gave companies a two-year free pass to launch high-risk AI without oversight. They’re hitting the brakes on regulation. At the exact same moment, Google’s paper hits the accelerator on an existential threat to the entire hundred-billion-dollar crypto ecosystem. The gap between what the technology can do and what our laws can handle isn't just growing—it’s accelerating.

The EU delay isn't just about giving companies more time. It’s an incentive. It tells every AI company with a high-risk system—like a hiring algorithm or a credit scoring model—to ship it NOW, before the rules kick in. Corporate lobbyists met with the EU commission on AI policy almost seventy percent of the time in 2025. This delay is what that kind of access buys you. Simultaneously, the quantum threat to Bitcoin is no longer a theoretical problem for the 2030s. Haseeb Qureshi at Dragonfly warns that every blockchain needs a transition plan before the end of this decade. But blockchains are decentralized. You can’t just push a software update like a bank can.

It requires massive, slow, contentious coordination. And Bitcoin has never been good at that. What this week reveals is a dangerous divergence. The code is moving at the speed of discovery, while the law is moving at the speed of lobbying. We are building world-breaking tools faster than we are building the guardrails to manage them. And right now, the builders are winning.