@drfeifei The measure of AI progress can’t be capability alone. It has to be whether that capability expands human agency, judgment, and flourishing.
Human-centered AI isn’t a constraint we add at the end. It’s an architectural choice from the beginning.
@Yoshua_Bengio@UN Safety must be architected into our systems from the start, not bolted on later. That means decentralization and layered safeguards—not just model eval at deployment, important as that is.
Dismissing these concerns as PR theater gives builders permission to look away.
What if "Can AI outsmart a human?" is the wrong question?
Better: can it outperform a human using AI?
AI wins where speed is the constraint. Most real work needs judgment, context, interaction.
Human + AI is the real benchmark. That's the centaur.
https://t.co/pzJ5aOQbMx
2026 Human-AI collaboration, meet 1983 automation control systems:
"If the human operator is not involved in on-line control, [they] will not have detailed knowledge of the current state of the system."
— Lisanne Bainbridge, Ironies of Automation, 1983
Oh, the irony.
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.
Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.
You can read the full post here: https://t.co/OGyPb7yaYt
@MsPackyetti I voted but didn’t get as involved in helping others vote bc polls said Trump couldn’t win. This year I am going to vote again but I also want to get involved in helping OTHERS vote. Ending the current administration is more important than any vote I have made in the past.
“Leadership is all about balance. You have to be able to take the high road and understand different view points to solve the world’s toughest challenges.” -keynote from Aicha Evans, CEO of Zoox #GHC19
“It’s ok to feel small. Sometimes I still do. But together we can build something bigger than all of us.” -keynote from Dr. Fei-Fei Li, Co-Director of Stanford’s Human Centered AI Institute #GHC19
Long badge line full of women engineers here at #GHC19. It’s emotional to see the culture of software evolving, but also to think about what it has historically looked like. I love all you ladies and I’m so glad we are all here working it and supporting each other. Keep it up!