The Next Phase of Artificial Intelligence
Why It Matters
Physical AI promises to transform robotics, drug discovery, and personalized medicine, creating a new market frontier that will dominate tech investment and product strategy over the next five years.
Key Takeaways
- •Physical AI and world models will dominate by 2026.
- •New multimodal, multiscale architectures needed for biology and robotics.
- •$1B+ funding fuels startups targeting physical AI over traditional data centers.
- •European hubs, especially Paris, provide deep talent for AI‑biology convergence.
- •Consumer‑grade physical AI expected within 3‑5 years, B2B now emerging.
Summary
The panel discussed the "next phase" of artificial intelligence, emphasizing a shift from language‑centric models to physical AI that can understand and act in the real world. Speakers highlighted world‑model research, multimodal and multiscale architectures, and the growing need for AI that operates on sensor, video, and biological data rather than discrete text tokens.
Yann LeCun projected 2026 as the breakthrough year for world models, noting a surge in academic papers and venture capital pouring over $1 billion into startups such as AMA Labs. JP Morrison stressed that traditional data‑center spending is giving way to specialized compute for robotics and drug discovery, where existing large‑language‑model techniques fall short because of infinite prediction spaces.
Concrete examples underscored the momentum: AI‑driven protein‑design tools are now deployed across biotech firms, and a new “wild model” can infer molecular activity from biopsy images, accelerating personalized medicine. The speakers also praised Paris’s deep talent pool in computer science, applied mathematics, and life sciences as a competitive alternative to Silicon Valley.
The implications are clear: B2B applications of physical AI will materialize within a year, while consumer‑grade systems—such as domestic robots—are projected to arrive in three to five years. Investors are repositioning toward this frontier, and companies that master multimodal, multiscale learning will shape the next industrial and healthcare revolutions.
Comments
Want to join the conversation?
Loading comments...