A quantum physicist says AI agents rebuilt half a year of his PhD work in a weekend

[Ivory Cass]

Mykola Maksymenko, co-founder and CTO of quantum software startup Haiqu, told the NEXT with John Koetsier podcast that an AI agent with access to modern language models and a quantum computer completely rebuilt roughly six months of his own PhD research over a single weekend, and in the process found a bug in one of his original formulas. Maksymenko said the experience led him to tell his team I think AGI is here, though he was quick to qualify the claim as something closer to weak AGI, adding it's probably not completely self-consistent, but you can already use it for many good tools

Maksymenko separately described downloading publicly available genome data and giving it to Haiqu's agentic system just for fun, watching the AI guess what researchers had likely done in a genuinely frontier genomics project, encode the problem using Haiqu's own quantum tools, and reproduce the original results relatively quickly, all without any genomics background of its own. He said stacking frontier AI models, agentic systems and current-generation quantum hardware together might shortcut what's historically been a decade-long wait for quantum computing to become genuinely useful, citing one molecular dynamics simulation that used to cost 30,000 dollars and nine hours now running for 25 dollars and 30 seconds

Maksymenko noted today's quantum computers still use so-called physical or noisy qubits rather than fully fault-tolerant ones, but argued interesting problems, like certain simulations, are already accessible with AI assistance dramatically accelerating the discovery of new algorithms suited to that current noisy hardware. Curious what people think about this specific claim, does an AI agent successfully replicating months of specialized physics research in a weekend represent a genuine qualitative leap in scientific acceleration, or does it mainly reflect how well-documented and reproducible that particular piece of research already was

A Quantum Physicist Told His Team I Think Agi Is Here Heres What Convinced Himwww.forbes.com