Just kidding. Recht might be even more suspicious of these things than we are. What has statistics ever done for us, really? When was the last time YOU ran a clinical trial after all, huh? HUH? After Ben Chugg defends his life decision to do a PhD in statistics, we talk AI, cults, philosophy, Paul Meehl, and discuss Ben Recht's forthcoming book, The Irrational Decision.

Check out Ben's blog, website, and his story about machine learning.

We discuss

• Ben Recht's theory of blogging
• Why is Berkeley the epicenter of AI doom?
  
• Where the word "robot" came from
• Is Bayesian reasoning responsible for AI doom?
• Paul Meehl and his contributions to science
• Ben Recht's bureaucratic theory of statistics
• What on earth is null hypothesis testing?
• What is the point of statistics?
• "Sweet spots" and "small worlds"
• Does science proceed by Popperian means?
• Can Popper get around the Duhem-Quine problem?
  

Errata

The z-score for the Pfizer trial was 20, not 12!

References

• Argmin, Ben Recht's blog
• David Freedman, UC Berkeley
• Paul Meehl's online course
• Theoretical Risks and Tabular Asterisks: Sir Karl, Sir Ronald, and the Slow Progress of Soft Psychology, Paul Meehl's 1978 paper.
• Clinical versus statistical prediction: A theoretical analysis and a review of the evidence, by Meehl
• On the near impossibility of estimating the returns to advertising
• A Bureaucratic Theory of Statistics by Recht
• The new riddle of induction by Goodman
• Announcing the Irrational Decision
• Patterns, Predictions, and Actions, textbook by Ben Recht and Moritz Hardt

Socials

• Follow us on Twitter at @BeenWrekt, @IncrementsPod, @BennyChugg, @VadenMasrani
• Come join our discord server! DM us on twitter or send us an email to get a supersecret link
• Become a patreon subscriber here. Or give us one-time cash donations to help cover our lack of cash donations here.
• Click dem like buttons on youtube

What's Berkeley's next cult? Send your guess over to incrementspodcast@gmail.com

Special Guest: Ben Recht.

Support Increments

Listener Note:

We strongly recommend watching the video for this one, available both on youtube and spotify:
- https://www.youtube.com/@incrementspod
- https://open.spotify.com/show/1gKKSP5HKT4Nk3i0y4UseB

We discuss

• The incentives of superforecasters
• Arguments by authority
• Whether superintelligence is right around the corner
• The difference between model size and data
• Are we running out of high quality data?
• Does training on synthetic data work?
• The assumptions behind the AGI claims
• The pitfalls of reasoning from trends

References

• Michael I Jordan
• Neil Lawrence
  
• [Important technical paper from Jordan pushing back on Doomerism](A Collectivist, Economic Perspective on AI)
• Jordan article talking about dangers of using AlphaFold data
• Nature paper showing you can't use synthetic data to train bigger models
• Paper estimating of when training data will run out (Coincidentally enough, sometime between 2027-2028)

Socials

• Follow us on Twitter at @IncrementsPod, @BennyChugg, @VadenMasrani
• Come join our discord server! DM us on twitter or send us an email to get a supersecret link
• Become a patreon subscriber here. Or give us one-time cash donations to help cover our lack of cash donations here.
• Click dem like buttons on youtube

But how many cubes until we get to AGI though? Send a few of your cubes over to incrementspodcast@gmail.com

Episode header image from here.

Support Increments

We discuss:

• How well does math fit reality? Is that surprising?
• Should artificial general intelligence (AGI) be considered "a person"?
• How could AI possibly "go rogue?"
• Can we know if current AI systems are being creative?
• Is misplaced AI fear hampering progress?

References:

• The Unreasonable effectiveness of mathematics
• Prohibition on autonomous weapons letter
• Google employee conversation with chat bot
• Gary marcus on the Turing test
• Melanie Mitchell essay.
• Did MIRI give up? Their (half-sarcastic?) death with dignity strategy
• Kerry Vaughan on slowing down AGI development.

Contact us

• Follow us on Twitter at @IncrementsPod, @BennyChugg, @VadenMasrani
• Check us out on youtube at https://www.youtube.com/channel/UC_4wZzQyoW4s4ZuE4FY9DQQ
• Come join our discord server! DM us on twitter or send us an email to get a supersecret link

Which prompt would you send to GPT-3 in order to end the world? Tell us before you're turned into a paperclip over at incrementspodcast@gmail.com

Support Increments