This post is probably hazardous for one type of person in one particular growth stage, and necessary for people in a different growth stage, and I don't really know how to tell the difference in advance.

If you read it and feel like it kinda wrecked you send me a DM. I'll try to help bandage it.

One of my favorite stories growing up was Star Wars: Traitor, by Matthew Stover.

The book is short, if you want to read it. Spoilers follow. (I took a look at it again recently and I think it didn't obviously hold up as real adult fiction, although quite good if you haven't yet had your mind blown that many times)

One anecdote from the story has stayed with me and permeates my worldview.

The story begins with "Jacen Solo has been captured, and is being tortured."

He is being [...]

The original text contained 3 footnotes which were omitted from this narration.

---

First published:

Source:

Narrated by TYPE III AUDIO.

The most likely person to write On the Edge was Nate Silver.

Grok thinks the next most likely was Michael Lewis, followed by a number of other writers of popular books regarding people thinking different.

I see why Grok would say that, but it is wrong.

The next most likely person was Zvi Mowshowitz.

I haven’t written a book for this type of audience, a kind of smarter business-book, but that seems eminently within my potential range.

On the Edge is a book about those living On The Edge, the collection of people who take risk and think probabilistically and about expected value. It centrally covers poker, sports betting, casinos, Silicon Valley, venture capital, Sam Bankman-Fried, effective altruism, AI and existential risk.

Collectively, Nate Silver calls this cultural orientation The River.

It is contrasted with The Village, which comprises roughly the mainstream [...]

---

Outline:

(02:53) Overview

(07:56) Introduction: The River

(14:21) Nate Silver Comes Home to The River

(18:29) Nate (In General) Makes One Critical Mistake

(18:57) The Village Idiots

(22:22) Alone in the Wilderness

(25:46) Why the River Hates the Village

(29:57) Nate Silver's History of River Versus Village

(38:46) Spending Time at Airports

(41:14) The Coin Flip

(42:15) The Other Risk Takers

(49:45) Aside on Covid

(50:46) What is a Contrarian?

(53:49) Prediction Market Smackdown

(56:53) Checkpoint One

The original text contained 1 image which was described by AI.

---

First published:

Source:

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

Testing an LM system for dangerous capabilities is crucial for assessing its risks.

Summary of best practices

Best practices for labs evaluating LM systems for dangerous capabilities:

• Publish results
• Publish questions/tasks/methodology (unless that's dangerous, e.g. CBRN evals; if so, just publish a small subset and offer to share more information with other labs, government, and relevant auditors)
• Do good elicitation and publish details (or at least demonstrate that your elicitation is good):
  • General finetuning (for "instruction following, tool use, and general agency" and maybe capabilities in the relevant area)
  • Helpful-only; no inference-time mitigations
  • Scaffolding, prompting, chain of thought
    • The lab should mention some details so that observers can understand how powerful and optimized the scaffolding is. Open-sourcing scaffolding or sharing techniques is supererogatory. If the lab does not share its scaffolding, it should show that the scaffolding is effective by running the same model with the [...]

---

Outline:

(00:11) Summary of best practices

(04:26) How labs are doing

(06:40) Appendix: misc notes on best practices

(06:49) Appendix: misc notes on particular evals

(12:08) Appendix: reading list

The original text contained 2 footnotes which were omitted from this narration.

---

First published:

Source:

Narrated by TYPE III AUDIO.

A common claim among e/accs is that, since the solar system is big, Earth will be left alone by superintelligences. A simple rejoinder is that just because Bernard Arnault has $170 billion, does not mean that he'll give you $77.18.

Earth subtends only 4.54e-10 = 0.0000000454% of the angular area around the Sun, according to GPT-o1.[1]

Asking an ASI to leave a hole in a Dyson Shell, so that Earth could get some sunlight not transformed to infrared, would cost It 4.5e-10 of Its income.

This is like asking Bernard Arnalt to send you $77.18 of his $170 billion of wealth.

In real life, Arnalt says no.

But wouldn't humanity be able to trade with ASIs, and pay Them to give us sunlight? This is like planning to get $77 from Bernard Arnalt by selling him an Oreo cookie.

To extract $77 from Arnalt, it's not a sufficient [...]

The original text contained 1 footnote which was omitted from this narration.

---

First published:

Source:

Narrated by TYPE III AUDIO.

In 2014 I got on SSRIs the first time, and they were amazing. I wrote online about how I suddenly had energy to do things, could concentrate on stuff, and generally just felt better and happier.

I now got a message from someone who’d found my writings and was wondering what my experience with antidepressants was now, 10 years later.

I wrote this reply to them, and thought I might as well share it with others:

—

Hi, that was indeed me!

I was on SSRIs for about a year after writing that comment, after that it felt like they started losing some of their effect but I also thought I felt better for other reasons, so I stopped using them. Then I was off them for about a year or two and started feeling bad again, so I got back on them. They [...]

---

First published:

Source:

Narrated by TYPE III AUDIO.

Hi all! I'm creating a deep dive podcast episode about (the original) Leverage Research (for my podcast, Clearer Thinking). If you worked at (or currently work at) Leverage, or you have had experiences connected to Leverage, and you would like to either potentially:

(1) do an interview for the podcast

(2) talk to me totally off the record about your experiences (i.e., with your privacy totally protected)

Please reach out privately by using my contact form here. Or you can email our podcast (but multiple team members who work on the podcast will see it) here:

I have known many people involved in Leverage over the years (from its founding to the present day), and so I have heard a great deal of information about it from others. But I've never been a member of it, and there are still many things I don't know about it.

The [...]

The original text contained 1 image which was described by AI.

---

First published:

Source:

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

This essay was written by @Metacelsus.

In March 2023, a research team at Osaka University announced an extraordinary achievement: the production of viable eggs using cells taken from male mice.

Building on years of research into stem cells and reproductive development,1 the researchers, led by Professor Katsuhiko Hayashi, developed a method to induce male stem cells to lose their Y chromosome and duplicate the X chromosome, thus becoming female. From there, researchers took the female cells and grew eggs from them using an elaborate protocol they had initially published in 2016.

From the paper and ensuing media fanfare, a question emerged: When will this technology become available for people?

For nearly all of human history, there was only one way to make a baby. When a man and a woman love each other very much . . . well, you know the rest. But for many people, the usual [...]

---

Outline:

(03:48) How Eggs and Sperm Get Made

(18:01) Reaching the Clinic

(25:42) Costly Prospects

The original text contained 5 images which were described by AI.

---

First published:

Source:

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

I’ve been fascinated recently by historical linguistics, and in particular reconstructed proto-languages such as Proto-Indo-European.

But what even is a “proto-language”?

Well, if we go for the literal operationalization, Wikipedia defines a proto-language as:

a postulated ancestral language from which a number of attested languages are believed to have descended by evolution, forming a language family.

[…]

It is by definition a linguistic reconstruction formulated by applying the comparative method to a group of languages featuring similar characteristics.

So a proto-language is supposed to be the closest common ancestor of a language family (or subfamily), and it is reconstructed using the comparative method (which as the name suggests, works by comparing the different daughter languages to reconstruct (mostly) sound structure (phonology) and word structure (morphology).

But they’re a weird kind of thing.

First, proto-languages are not attested. This means that we have no example of writing in any proto-language. [...]

The original text contained 8 footnotes which were omitted from this narration.

---

First published:

Source:

Narrated by TYPE III AUDIO.

expectation calibrator: freeform draft, posting partly to practice lowering my own excessive standards.

So, a few months back, I finally got around to reading Nick Bostrom's Superintelligence, a major player in the popularization of AI safety concerns. Lots of the argument was stuff I'd internalized a long time ago, reading the sequences and even living with AI alignment researchers for a few months. But I wanted to see the circa-2015 risk-case laid out as thorougly as possible, partly as an excusion to learn more about my own subculture, and partly to see if I'd missed anything crucial to why all of us consider future advanced AI systems so dangerous.

But for the most part, the main effect of reading that book was actually to finally shatter a certain story I'd been telling myself about the AI doom scenario, the cracks in which had been starting to show for several [...]

The original text contained 2 footnotes which were omitted from this narration.

---

First published:

Source:

Narrated by TYPE III AUDIO.

This is a guest post by my student Ruiqi Zhong, who has some very exciting work defining new families of statistical models that can take natural language explanations as parameters. The motivation is that existing statistical models are bad at explaining structured data. To address this problem, we agument these models with natural language parameters, which can represent interpretable abstract features and be learned automatically.

Imagine the following scenario: It is the year 3024. We are historians trying to understand what happened between 2016 and 2024, by looking at how Twitter topics changed across that time period. We are given a dataset of user-posted images sorted by time, $x_1$, $x_2$ ... $x_T$, and our goal is to find trends in this dataset to help interpret what happened. If we successfully achieve our goal, we would discover, for instance, (1) a recurring spike of images depicting athletes every four [...]

---

Outline:

(04:21) Warm-up Example: Logistic Regression with Natural Language Parameters

(07:12) Formalizing Regression with Natural Language Parameters.

(10:47) A Modeling Language for Specifying More Complex Models

(14:20) Example Applications:

(14:49) Application 1: monitoring trends in LLM usage via time series modeling

(16:01) Application 2: taxonomizing product reviews via clustering

(18:05) Application 3: Explaining memorable visual features via Classification.

(20:12) Conclusion

The original text contained 5 footnotes which were omitted from this narration.

The original text contained 2 images which were described by AI.

---

First published:

Source:

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.