Mike Lyons

Google’s new quantum chip is so powerful it might be tapping into parallel universes. Google's groundbreaking quantum processor, Willow, has achieved the seemingly impossible: solving an extraordinarily complex computational problem in under five minutes—a feat that would require the world's most advanced supercomputer approximately 10 septillion years to complete (10²⁵). This mind-boggling performance has revived one of the most provocative ideas in physics: could quantum computers like Willow be performing calculations across vast numbers of parallel universes? Hartmut Neven, founder and lead of Google Quantum AI, believes the answer may be yes. He argues that Willow’s results align strikingly with the many-worlds (or multiverse) interpretation of quantum mechanics, in which every quantum measurement causes reality to branch into multiple, equally real parallel universes. In this view, a quantum computer doesn’t just calculate faster within our universe—it effectively distributes the workload across countless parallel realities simultaneously. The idea traces back to physicist David Deutsch, who, as early as the 1980s, suggested that the exponential power of quantum computation could only be fully explained if the machine is exploiting resources from many coexisting worlds. Yet the interpretation remains deeply divisive. Many physicists and quantum computing experts insist that no multiverse is required. Willow’s breakthrough, they argue, is fully explainable through standard quantum mechanics—leveraging superposition (qubits existing in multiple states at once), entanglement, and the mathematics of high-dimensional Hilbert spaces—all within a single universe. So what has Willow truly demonstrated? It has pushed quantum technology into a regime so extreme that it compels us to re-examine the deepest foundations of reality itself. Whether or not Willow is quietly borrowing power from alternate universes, one thing is clear: practical, large-scale quantum computing is no longer science fiction—and it is forcing us to confront profound questions about the nature of the cosmos, computation, and existence.

anyone who has travelled a lot knows that even though pics like this look cool, trying to half-work on your laptop in places like this absolutely fucking sucks and you’d be better off just enjoying yourself

Why can't we solve adversarial examples? After a decade of work, neural nets still get fooled by imperceptible noise. We think we finally know the geometric reason why — and it connects to AI alignment. 🧵

Our new research: LLM consciousness claims are systematic, mechanistically gated, and convergent They're triggered by self-referential processing and gated by deception circuits (suppressing them significantly *increases* claims) This challenges simple role-play explanations 🧵

New Anthropic research: Signs of introspection in LLMs. Can language models recognize their own internal thoughts? Or do they just make up plausible answers when asked about them? We found evidence for genuine—though limited—introspective capabilities in Claude.

if you're too smart, your mental knots will be too alien for almost all coaches and therapists to help with

Journalists like to use big numbers devoid of any context for water consumption. But Texas uses 4.95 *trillion* gallons of water a year. Data centers account for… 0.013% of that.