Quantum Meets AI | Alan Ho (CEO, Qolab)

Listen to full episode:

Summary: Anant and Ed start the episode discussing OpenClaw and the local-compute surge it's driving, plus the growing tendency for users to get attached to specific model personalities. Alan Ho (CEO, Qolab) then joins, drawing on his work on Google's quantum supremacy experiment to argue quantum computing is AI's next real inflection point: a calculation that would take classical supercomputers thousands of years ran in seconds on a quantum machine, though he's clear the tech is nowhere near cracking Bitcoin's cryptography yet. The conversation closes on economics — foundation model training costs climbing into the billions, forcing massive cluster buildouts at Google and Amazon — with Alan's bet that quantum computing unlocks the next generation of models, starting with drug discovery and autonomous vehicles.

Chapters:

00:00 - Opening banter and why quantum is back on the agenda

02:00 - OpenClaw, agentic workflows, and the Super Bowl ad moment for AI

09:23 - Introducing Alan Ho and his background in quantum supremacy at Google

16:29 - Why the next generation of AI may be quantum AI

20:51 - Quantum-centric supercomputing as the next foundation model trainer

25:04 - The real constraint is power, not just cost

28:54 - Quantum as a weapon in the foundation model race

33:16 - Noisy quantum applications, physical vs. logical qubits, and gate quality

42:57 - Quantum-safe cryptography and the risk to Bitcoin wallets

47:13 - Which countries are closest to fault-tolerant quantum

51:39 - AlphaQubit and AI helping with qubit error correction

54:31 - Reflections on timelines and next episode tease

Sound Bites:

“My hypothesis is that the next generation of AI, which currently today is around generative AI foundation models, is likely to be quantum AI.”

“The biggest trade-off is going to be actually in power.”

“Quantum computing is most likely to be a weapon in the foundation model battle.”