If you want the 60 second version of this post that just tells you what to do, click here to skip to the summary.

Every now and then I stumble down a YouTube rabbit hole and end up learning way more than I ever intended about a particular topic. A few months ago I noticed that I was starting to get permanent smile lines etched around my mouth and started wondering whether there was something I could do to fix them.

An ex of mine was a skincare addict and had perhaps the nicest skin of anyone I have ever met. I texted her to ask for advice and she recommended I use a retinoid.

Since I didn't know what those were or how they worked, I watched a YouTube video. Thus began my 3 month journey down the rabbit hole of skin care products and reviews and progress [...]

---

Outline:

(01:04) What are retinoids?

(04:19) Ok, Im sold. Where do I get a retinoid?

(05:18) How to use retinoids

(06:10) What else should I know about?

(06:59) Use a cleanser

(07:29) Use moisturizer

(07:49) Use Suncreen

(08:42) TL;DR Version

The original text contained 8 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.

For many products, we face a choice of who to hold liable for harms that would not have occurred if not for the existence of the product. For instance, if a person uses a gun in a school shooting that kills a dozen people, there are many legal persons who in principle could be held liable for the harm:

Which one of these is the best? I'll offer a brief and elementary economic analysis of how this decision should be made in this post.

The important concepts from economic theory to understand here are Coasean bargaining and the problem of the judgment-proof defendant.

Coasean bargaining

Let's start with Coaesean bargaining: in short, this idea says that regardless of [...]

---

Outline:

(00:49) Coasean bargaining

(02:09) The judgment-proof defendant

(04:20) Transaction costs and economies of scale

(05:23) Summary and implications for AI

---

First published:

Source:

Narrated by TYPE III AUDIO.

Meta: I recently entered the dating market, so naturally I have lots of random thoughts on the subject which you all get to suffer through for a while. Your usual diet of dry math and agency theory will resume shortly.

Obviously the phrase “true love” has been so thoroughly overdone in so much fiction as to lose all substantive meaning. That's what happens when we leave important conceptual work to would-be poets. We’re here to reclaim the term, because there's a useful concept which is very naturally described by the words “true” and “love”.

You know that thing where, when you’re smitten by someone, they seem more awesome than they really are? Your brain plays up all the great things about them, and plays down all the bad things, and makes up stories about how great they are in other ways too? And then you get even more smitten [...]

---

First published:

Source:

Narrated by TYPE III AUDIO.

I don't usually post stories on LessWrong so I'm curious to see how this is received.

The first time we spoke, you asked me some questions that felt really invasive. I didn’t want that to happen again, so I avoided you the entire following year.

So when you said “Hi” at a party and suggested catching up, I hesitated. But curiosity won out.

You still asked probing questions like “Why did you quit your job?” and “What did you think of your manager? I hear they don't have a great social reputation.”

These weren't questions I wanted to answer. But this time, something was different. Not you—me.

In the past, I would have felt forced to answer your questions. But I’m sure you can remember how I responded when we spoke again: “Mm, I don’t want to answer that question”, “I don’t want to gossip” [...]

---

Outline:

(02:46) Addendum: Beliefs I have

(02:51) Emotional security is the absence of insecurities

(03:34) It's about present predictions, not past trauma

---

First published:

Source:

Narrated by TYPE III AUDIO.

I recently argued for Cohabitive Games, games that are designed for practicing negotiation, or for developing intuitions for applied cooperative bargaining, which is one of the names of preference aggregation.

I offered the assets for P1, my prototype, to whoever was interested. About 20 people asked for them. I told most of them that I wanted to polish the assets up a little bit first, and said that it would be done in about a month. But a little bit of polish turned into a lot, and other things got in the way, and a month turned into ten. I'm genuinely a little bit dismayed about how long it took, however:

P1, now Optimal Weave 0.1, is a lot nicer now.
I think it hangs together pretty well.

You can get it here: Optimal Weave 0.1. (I'm not taking any profit, and will only start to if it runs [...]

---

Outline:

(02:36) What Happened

(02:43) I refined the game a fair bit

(03:47) I found a nicer low-volume prototyping service

(04:47) Automating the pipeline

(07:15) Final Autumn Together

(09:24) Final Autumn: Rules:

(11:22) Now is the time for everyone to join in.

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.

Did you know that an autoimmune disease now affects one in five Americans? Of these, 80% are women; autoimmune disease is the fifth leading cause of death in women under 65. Most autoimmune diseases are being diagnosed in increasing numbers, but it is still unclear if all this variance is due to higher awareness, medical literacy, or better diagnosis techniques.

The effects of autoimmune diseases can be devastating. As a person's immune system attacks their body instead of microbes or cancerous cells, they can experience chronic fatigue, chronic pain, drug dependency, depression, and social isolation. These symptoms annihilate mental health, wreck promising careers, destroy lives, and, often, financially ruin families. For too many, these illnesses result in early death.

The Missing Friends

The Old Friends Hypothesis suggests that early and regular exposure to harmless microorganisms—"old friends" present throughout human evolution and recognized by the human immune system—train the immune [...]

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.

The agent-like structure problem is a question about how agents in the world are structured. I think rationalists generally have an intuition that the answer looks something like the following:

• We assume the world follows some evolution law, e.g. maybe deterministically like _x_{n+1} = f(x_n)_, or maybe something stochastic. The intuition being that these are fairly general models of the world, so they should be able to capture whatever there is to capture. _x_ here has some geometric structure, and we want to talk about areas of this geometric structure where there are agents. 
• An agent is characterized by a Markov blanket in the world that has informational input/output channels for the agent to get information to observe the world and send out information to act on it, intuitively because input/output channels are the most general way to model a relationship between two systems, and to embed one system within another [...]

---

Outline:

(02:47) Radical computationalism is killed by information overload

(04:59) Energy-orientation solves information overload

(06:35) ... kinda

---

First published:

Source:

Narrated by TYPE III AUDIO.

Short version

It has always felt like a gift that people read things I write, and I enjoyed reciprocating that by sharing my writing for free. Until recently I had enough income to pull that off and still cover material needs. That is no longer true, and while it is not an urgent problem, I would like to solve it while that is still true. However it's important to me to keep this blog free. To square this, I’m asking for your support. If you value this blog and can comfortably do so, please consider sending some of that value back my way via Patreon or Paypal.

Long version

I love my work and this blog. Clients paying me to learn interesting things that they immediately use to help people would be amazing on its own, but then grantmakers pay me to do more speculative EV [...]

---

Outline:

(00:03) Short version

(00:42) Long version

(02:52) How to give me money

(04:12) Current and Potential Projects

(06:21) Counterarguments

(06:24) Fungibility

(08:04) Low-urgency

(09:10) Why I want this blog to stay free

(10:10) Conclusion

---

First published:

Source:

Narrated by TYPE III AUDIO.

There's an admirable LW post with (currently) zero net upvotes titled Goodhart's Law and Emotions where a relatively new user re-invents concepts of related to super-stimuli. In the comments, noggin-scratch explains in more detail:

The technical meaning is a stimulus that produces a stronger response than the stimulus for which that response originally evolved.

So for example a candy bar having a carefully engineered combination of sugar, fat, salt, and flavour in proportions that make it more appetising than any naturally occurring food. Or outrage-baiting infotainment "news" capturing attention more effectively than anything that one villager could have said to another about important recent events.

In my opinion, there's a danger that arises when applying the dictum to know thyself, where one can do this so successfully that one begins to perceive the logical structure of the parts of yourself that generate subjectively accessible emotional feedback signals.

In the [...]

---

First published:

Source:

Narrated by TYPE III AUDIO.

TL;DR: We demonstrate that the decoder directions of GPT-2 SAEs are highly structured by finding a historical date direction onto which projecting non-date related features lets us read off their historical time period by comparison to year features.

Calendar years are linear: there are as many years between 2000 and 2024, as there are between 1800 and 1824. Linear probes can be used to predict years of particular events from the activations of language models. Since calendar years are linear, one might think the same of other time-based features such as weekday features, however weekday activations in sparse autoencoders (SAEs) were recently found to be arranged in a circular configuration in their top principal components. Inspired by this, we looked into weekdays, months, and most interestingly calendar years from the perspective of SAE feature decoder similarity.

For [...]

---

Outline:

(01:35) Features at different scales

(05:19) Interpreting Principal Components

(07:34) Finding Directions

---

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.