2026 Predictions: Physical AI, Quantum Computing & the Future of Work

Von Tobel bets infrastructure will learn to fix itself

Alexa von Tobel's 2026 predictions point the same direction. Software is moving from reactive to predictive, and the physical world is where it shows up first, in water pipes, power lines and self-driving cars. Quantum computing is the exception, a decade-out bet she tracks by counting qubits rather than deploying them. The payoff she describes is time, as machines absorb routine work and humans keep strategic creativity, complex problem solving and better jobs.

Takeaways

• Physical AI closes the data-to-action loop. AI leaves the lab to reshape manufacturing, utilities, climate monitoring and logistics. Smart machines become adaptive and self-correcting, moving infrastructure from reactive to predictive. Von Tobel calls it the dawn of autonomous infrastructure, not robotics 2.0.
• Quantum computing remains years from deployment. Von Tobel spent a year and a half studying quantum and predicts more than 10,000 qubits between now and 2027, after which experiments and material improvements begin. She does not expect it today, probably not in five years, potentially in ten.
• AI takes jobs people should not have. Von Tobel rejects the claim that AI will eliminate 100m jobs. She expects it to handle routine, mundane work such as checking subway traps for rodents, while a human plus AI makes people far more productive so existing jobs get better.
• Humans keep strategic creativity and complex problem solving. AI already summarizes meetings and notes and says what comes next, freeing thinking for the future. Von Tobel argues humans will probably stay better at strategic creativity and complex, multidimensional problem solving because of the inputs and context they bring.
• Autonomous vehicles are her fastest-wanted prediction. Riding in a Waymo in San Francisco is quiet and private. She wants Waymo trusted in American suburbs to drive children to sports while parents watch on their phones. Suburban mothers spend 20 to 30 hours a week driving.
• Freeing low-value time is the through-line. Duckbill AI, which Inspired Capital backed, plans events, buys gifts, manages logistics, books doctor's appointments and finds HVAC companies for her, and she says she has not done a mundane task in a year. Autonomy gives people more creative and more leisure time.

How sensing fixes leaky pipes

1. United States utilities lose more than 6bn gallons of treated water a day, enough to supply 15m homes, so sensing devices are sent into pipes to find leaks proactively and fix them.
2. The devices stay in the pipes to monitor for another leak or a breakage.
3. The same proactive approach extends to other infrastructure, so a leaning telephone pole is known about and fixed before it falls and starts a fire.