Loading articles…
The complete subject coverage could not load. Published articles remain below.
Dileep George, who left DeepMind to lead Astera's brain-inspired AI effort, argues scaling language models works without the principles of intelligence, leaving perhaps a dozen discoveries to make.
Warp CTO Manohar, until weeks ago a Meta VP, argues frontier AI needs verifiable physical feedback loops to keep improving, while conceding closed labs will likely keep leading on capability.
DeepMind's Demis Hassabis argues scaling still pays but frontier advantage is shifting toward algorithmic invention, while real-world bottlenecks like clinical trials and international governance determine how AGI's benefits arrive.
Geoff Hinton argues chatbots already understand and will outthink humans, yet concedes his own timing forecasts missed and that nobody has a real safety solution.
Sergey Levine of Physical Intelligence on how robot foundation models evolved from pi 0 to pi 0.7, why generality beats specialization, and where robot data will come from.
MIT physicist Max Tegmark, talking with Brian Keating, covers OpenAI's partial Navier-Stokes result, the Hugging Face agent breakout, and why safety standards and formally verified code could lower P(doom).
Yann LeCun, founder of AMI Labs and ex-Meta chief AI scientist, on JEPA world models, why he left Meta, and why open models will win.