Show Notes
This year, two powerful AI CEOs said they want to stop building it and will… if everyone else agrees to stop too. Yet they race ahead with stock market valuations premised on making millions of workers redundant. If you wanted to stoke a popular revolt against AI, you couldn't design a better plan.
In this episode, Stuart Russell, the founder of UC Berkeley's Center for Human-Compatible AI, describes a $15 quadrillion prize, a singularity from the future sucking nearly all the money on Earth into its depths. If such a dangerous system gets loose, the only answer is the size of the problem: we would probably have to shut down the internet.
YouTube: https://www.youtube.com/@Futurologypod
Full playlist: https://www.youtube.com/playlist?list=PLwkiuf1eRbAAHP4p--2IFJbn0CIdaZMva
Apple Podcasts: https://podcasts.apple.com/us/podcast/futurology/id1821718921
Spotify: https://open.spotify.com/show/2I38HvHP6KlXrD5ysfygxk
Mentioned in this Episode:
Human Compatible: Artificial Intelligence and the Problem of Control — Stuart Russell (Book, 2019)
Artificial Intelligence: A Modern Approach — Stuart Russell and Peter Norvig (Textbook, first edition 1995)
Superintelligence: Paths, Dangers, Strategies — Nick Bostrom (Book, 2014)
Speculations Concerning the First Ultraintelligent Machine — I.J. Good (Essay, 1965)
Intelligent Machinery, A Heretical Theory — Alan Turing (Lecture, c. 1951)
Economic Possibilities for our Grandchildren — John Maynard Keynes (Essay, 1930)
The Diamond Age: Or, A Young Lady's Illustrated Primer — Neal Stephenson (Novel, 1995)
Dune — Frank Herbert (Novel, 1965)
Erewhon: Or, Over the Range — Samuel Butler (Novel, 1872)
Where to find Stuart Russell
Faculty page: https://people.eecs.berkeley.edu/~russell/
Center for Human-Compatible AI: https://humancompatible.ai/
Human Compatible (book): https://www.penguinrandomhouse.com/books/566677/human-compatible-by-stuart-russell/
Show ideas and feedback? Email: [email protected]
Learn more about the Berggruen Institute
Follow Futurology!
Instagram: / futurologypod
Twitter/X: / futurologypod
Tiktok: / futurologypod
Facebook: / berggrueninst
LinkedIn: / berggrueninst
Bluesky: / futurologypod
Credits
Executive Producers: Nicolas Berggruen, Nathan Gardels, Nils Gilman, Dawn Nakagawa, & Jason Hoch
Producers: Grant Slater, Alex Gardels, & Nathalia Ramos
Associate Producer: Elissa Mardiney
Theme Music: Marcus Bagala
Audio Engineer: Kyle Scott Wilson
Futurology is a production of Studio B and Wavland for the Berggruen Institute in Los Angeles, California.
More from Futurology
Hypocrisy, the old line goes, is the tribute vice pays to virtue. For eighty years, the powerful at least bothered to pay it, citing human rights, national security, and democracy as reasons for starting wars and undercutting liberties. Now the veneers have been stripped, and the next generation is not accepting business as usual.
Machines are starting to act like humans. Does that mean they’re conscious? The problem is we haven’t been able to prove how or why humans are conscious yet either. So why would we be able to offer any answers about AI?
Democracy counts people, not plants. Humans are only a sliver of life on this planet, but today, we're the only ones who get a vote. What if we gave rights to rivers and standing to trees?
Somewhere around the Enlightenment we decided that thinking was what made us human. We built our schools, careers, and sense of worth at the altar of intelligence. Now machines are passing us on that rubric, which leaves two possibilities: we were replaceable all along, or we picked the wrong standard to live by.
Relying on machines to coordinate our lives was just one step toward decentering humanity’s role in the world. Flights, banking, trade, food production... But as we start merging our actual decision-making with digital entities, exactly which pieces of humanity are we leaving behind? And how might we bring those pieces back?