Yuval Noah Harari and Fei-Fei Li discuss the challenges AI poses to society as the merger of biological knowledge and tech may enable hacking of humans at scale
Nicholas Thompson / Wired : Tweets: @alexjohnlondon , @drfeifei , and @wired Tweets: Alex John London / @alexjohnlondon : The interview captures a lot that I dislike in the discourse around AI: it sounds a sleepover conversation from eager high school students, complete with pseudo equation in Harari's opening bit to “explain what's going on”. http://www.wired.com/... Fei-Fei Li / @drfeifei : The full transcript of the interview by @nxthompson for me and @harari_yuval is out. It's a conversation about #AI and our shared future by a humanist and a technologist. @StanfordHAI We hope to see more such dialogue taking place. http://twitter.com/... @wired : “The easiest people to manipulate are the people who believe in free will, because they think they cannot be manipulated.” https://wired.trib.al/Ib14RkK
Context & Ripple Effects
Wired has been running its AI-and-society interview franchise for years — the 2016 Barack Obama and Joi Ito conversation set the template of pairing a political-philosophical voice with a technologist. This Harari–Fei-Fei Li exchange extends that format to the sharpest claim yet: that merging biological knowledge with computing could make human behavior manipulable at scale.
The interview lands inside an already-polarized risk discourse — Max Tegmark's existential-threat framing on one side, Yann LeCun's argument that AI danger is overhyped on the other — and drew immediate pushback from ethicist Alex John London, who called the exchange pseudo-profound. That reception matters as much as the claims themselves.
First-order effects
- Fei-Fei Li is positioned as the bridge figure between the doomer and skeptic camps — a Stanford HAI-affiliated technologist endorsing serious discussion of biological-scale manipulation while building commercial AI systems.
- The 'hacking humans at scale' thesis enters mainstream discourse through Wired's platform, forcing ethicists like London into public rebuttal rather than private critique.
Second-order effects
- The risk debate hardens into named camps — Tegmark-style existentialists versus LeCun-style skeptics — with figures like Fei-Fei Li courted by both sides for legitimacy.
- Claims about biotech-computing convergence put pressure on regulators already identified in later coverage as lagging, since no existing framework covers manipulation of humans at biological scale.
Third-order effects
- If the pattern holds, AI governance splits into two tracks — algorithmic safety and human-biological integrity — with the latter requiring coordination across bioethics bodies that today have little overlap with AI policy institutions.
The trend: Public AI-risk debate is consolidating around celebrity pairings of humanists and technologists, whose claims are increasingly contested by academic ethicists even as regulatory frameworks trail both.