Home Tim Ferriss Notes
Tim Ferriss · 2022-08-02 · 1h 35m

Will MacAskill of Effective Altruism Fame — The Value of Longtermism, AI, and How to Save the World

Philosopher Will MacAskill explains longtermism, existential risks from AI, pandemics and war, and why the vast future is worth protecting.

Will MacAskill of Effective Altruism Fame — The Value of Longtermism, AI, and How to Save the World
The guest

Will MacAskill — Associate professor of philosophy at Oxford, co-founder of Giving What We Can, the Centre for Effective Altruism and 80,000 Hours, and author of 'What We Owe the Future'.

The gist

Tim Ferriss talks with philosopher Will MacAskill about longtermism, the idea that protecting and improving the long-term future of humanity is a moral priority because almost everything we value lies ahead of us. They cover effective altruism's track record, the three biggest risks MacAskill sees in our lifetimes (advanced AI, engineered pandemics, and a third world war), and the concept of 'value lock-in' illustrated through historical examples. MacAskill argues for optimism grounded in the difference individuals can make rather than the scale of problems, and discusses concrete projects like far-UVC lighting and technical AI safety. The episode also covers personal topics including his productivity check-in system, his self-designed back-pain routine, and how he manages low mood with a 'trigger action plan.'

Big reveals

  • MacAskill says effective altruism has moved well over a billion dollars to effective causes; via the Against Malaria Foundation alone they've protected over 400 million people and statistically saved about 100,000 lives.
  • He ranks the most important events likely in our lifetimes as the development of human-level (or smarter) AI, catastrophic engineered pandemics far worse than COVID-19, and a third world war.
  • Over the next 10 years he's most concerned about AI; over 50 years both AI and war, estimating roughly a one-in-three chance of a third world war in our lifetime.
  • He argues the 2020s are more likely than any later decade to see truly transformative AI, because today's biggest AI systems use computing power comparable only to a honeybee's brain and human-level compute could arrive in roughly 10 years.
  • He lays out two worst-case AI scenarios: misaligned superhuman AI that takes control (humans becoming to AI what chimps are to humans), and aligned AI that concentrates power in a tiny number of hands enabling indefinite totalitarian 'value lock-in.'
  • He highlights far-UVC lighting as a project he's extremely excited about, arguing that installing it in lighting worldwide could make major progress toward never having another pandemic and even eradicating respiratory disease.
  • His top action items: take the Giving What We Can 10% pledge and donate (e.g., to the Long-Term Future Fund), learn more, then leverage or switch your career toward the most important problems.

Things worth remembering

  • A middle-class person in a rich country can save dozens, hundreds, or even thousands of lives over a lifetime through effective giving, per MacAskill.
  • Corporate cage-free campaigns the EA community helped fund mean hundreds of millions of hens are no longer in cages.
  • On the scale of a typical mammalian species lifespan, humanity today would be only about six months old, with over 99% of its potential life still ahead.
  • MacAskill describes ancient China's 'hundred schools of thought,' noting the Mohists resembled effective altruists and built a paramilitary group to defend besieged cities so that no war could be won.
  • After the Qin's legalism collapsed in 14 years, the Han Dynasty entrenched Confucianism as state ideology, which persisted for roughly 2,000 years.
  • Under the Khmer Rouge's Pol Pot in Cambodia, about 25% of the population was killed.
  • MacAskill credits a roughly 20-25% productivity boost during book-writing to 10-minute evening check-ins with an employee/productivity coach, capping caffeine at three espressos (about 180 mg).
  • He largely fixed his lower back pain with a self-designed bosu-ball workout, citing that cultures which squat to sit have lower rates of back pain.
  • He says he's now roughly 5 to 10 times happier than a decade ago and uses a 'trigger action plan' (going to the gym, meditating, zooming out to longer time horizons) to handle low mood.
  • He calls Our World in Data the single best source for a big-picture, data-grounded understanding of the world.

Recommended in this episode

Books, products and media the guest or host genuinely endorsed here — with the buy link.

Affiliate link — we may earn a commission at no extra cost to you.

RecommendedBook

The Precipice

Toby Ord

“one is the precipice by my colleague Toby Ord who I co-founded giving what we can with back in 2009 and it's on the topic of existential risks” — Will MacAskill 00:02:36
Find it on Amazon
RecommendedBook

The Secret of Our Success

Joseph Henrich

“that's Joe Heinrich's books secret of our success and the WEIRDest people in the world and Joe Henrich is a kind of quantitative anthropologist at Harvard” — Will MacAskill 00:03:38
Find it on Amazon
RecommendedBook

The WEIRDest People in the World

Joseph Henrich

“he has this other book the weirdest people in the world which is about the psychology in particular of weird people Western educated industrialized rich and Democratic” — Will MacAskill 00:05:43
Find it on Amazon
Guest’s ownBook

What We Owe the Future

Will MacAskill

“his new book is what we owe the future It Is blurbed by several guests of this podcast including neuroscientist and author Sam Harris” — Tim Ferriss 00:00:31
Find it on Amazon
RecommendedProduct

PSO-Rite

PSO-Rite (inferred)

“the device that was recommended to me a few times before I finally bit the bullet and got it was something called the PSO-Rite” — Tim Ferriss 00:35:01
Find it on Amazon
RecommendedMedia

In Our Time

BBC (inferred)

“for podcasts in our time hosted by Melvyn Bragg I believe discusses history philosophy science with leading academics” — Will MacAskill 01:31:39
Find it on Amazon
RecommendedMedia

80,000 Hours Podcast

80,000 Hours (inferred)

“and the 80,000 hours podcast would those be two you would still recommend yeah I would still strongly recommend them” — Will MacAskill 01:31:39
Find it on Amazon
RecommendedMedia

Hear This Idea

Finn Moorhouse and Luca Righetti (inferred)

“there's also another podcast he this idea by Finn Moorhouse and Luca Righetti I also particularly like” — Will MacAskill 01:32:11
Find it on Amazon
RecommendedMedia

Rationally Speaking

Julia Galef (inferred)

“rationally speaking by Julia Galef as well is like very good” — Will MacAskill 01:32:11
Find it on Amazon