I didn’t use to be, but now I’m part of the 2% of U.S. households without a television. With its near ubiquity, why reject this technology?

The Beginning of my Disillusionment

Neil Postman's book Amusing Ourselves to Death radically changed my perspective on television and its place in our culture. Here's one illuminating passage:

We are no longer fascinated or perplexed by [TV's] machinery. We do not tell stories of its wonders. We do not confine our TV sets to special rooms. We do not doubt the reality of what we see on TV [and] are largely unaware of the special angle of vision it affords. Even the question of how television affects us has receded into the background. The question itself may strike some of us as strange, as if one were to ask how having ears and eyes affects us. [In the 1960s], the question [...]

---

Outline:

(00:14) The Beginning of my Disillusionment

(01:47) A Man who was Crazy Enough

(03:15) A Mountain of Evidence

(07:48) I’m not an addict, am I?

(08:55) Struggling with Quitting

(10:36) A Quiet Place

---

First published:

Source:

Narrated by TYPE III AUDIO.

Introduction to a sequence on the statistical thermodynamics of some things and maybe eventually everything. This will make more sense if you have a basic grasp on quantum mechanics, but if you're willing to accept "energy comes in discrete units" as a premise then you should be mostly fine.

The title of this post has a double meaning:

1. Forget the thermodynamics you've learnt before, because statistical mechanics starts from information theory.
2. The main principle of doing things with statistical mechanics is can be summed up as follows:

Forget as much as possible, then find a way to forget some more.

Particle(s) in a Box

All of practical thermodynamics (chemistry, engines, etc.) relies on the same procedure, although you will rarely see it written like this:

1. Take systems which we know something about
2. Allow them to interact in a controlled way
3. Forget as much as possible
4. If we [...]

---

Outline:

(00:42) Particle(s) in a Box

(02:25) The Piston

(03:29) Maths of the Piston

(04:57) Conclusions

---

First published:

Source:

Narrated by TYPE III AUDIO.

This comes from a podcast called 18Forty, of which the main demographic of Orthodox Jews. Eliezer's sister (Hannah) came on and talked about her Sheva Brachos, which is essentially the marriage ceremony in Orthodox Judaism. People here have likely not seen it, and I thought it was quite funny, so here it is:

https://18forty.org/podcast/channah-cohen-the-crisis-of-experience/

David Bashevkin:

So I want to shift now and I want to talk about something that full disclosure, we recorded this once before and you had major hesitation for obvious reasons. It's very sensitive what we’re going to talk about right now, but really for something much broader, not just because it's a sensitive personal subject, but I think your hesitation has to do with what does this have to do with the subject at hand? And I hope that becomes clear, but one of the things that has always absolutely fascinated me about [...]

---

First published:

Source:

Narrated by TYPE III AUDIO.

Disclaimer: While I criticize several EA critics in this article, I am myself on the EA-skeptical side of things (especially on AI risk).

I am a proud critic of effective altruism, and in particular a critic of AI existential risk, but I have to admit that a lot of the critcism of EA is hostile, or lazy, and is extremely unlikely to convince a believer.

Take this recent Leif Weinar time article as an example. I liked a few of the object level critiques, but many of the points were twisted, and the overall point was hopelessly muddled (are they trying to say that voluntourism is the solution here?). As people have noted, the piece was needlessly hostile to EA (and incredibly hostile to Will Macaskill in particular). And he's far from the only prominent hater. Emille Torres views EA as a [...]

---

Outline:

(02:23) No door to door atheists

(04:52) What went wrong here?

(08:42) Motivation gaps in AI x-risk

(11:01) EA gap analysis

(15:14) Counter-motivations

(25:51) You can’t rely on ingroup criticism

(29:12) How to respond to motivation gaps

---

First published:

Source:

Linkpost URL:
https://titotal.substack.com/p/motivation-gaps-why-so-much-ea-criticism

Narrated by TYPE III AUDIO.

Concerns over AI safety and calls for government control over the technology are highly correlated but they should not be.

There are two major forms of AI risk: misuse and misalignment. Misuse risks come from humans using AIs as tools in dangerous ways. Misalignment risks arise if AIs take their own actions at the expense of human interests.

Governments are poor stewards for both types of risk. Misuse regulation is like the regulation of any other technology. There are reasonable rules that the government might set, but omission bias and incentives to protect small but well organized groups at the expense of everyone else will lead to lots of costly ones too. Misalignment regulation is not in the Overton window for any government. Governments do not have strong incentives to care about long term, global, costs or benefits and they do have strong incentives [...]

---

Outline:

(01:18) Default government incentives

(05:40) Negative Spillovers

---

First published:

Source:

Linkpost URL:
https://www.maximum-progress.com/p/ai-regulation-is-unsafe

Narrated by TYPE III AUDIO.

It was all quiet. Then it wasn’t.

Note the timestamps on both of these.

Dwarkesh Patel did a podcast with Mark Zuckerberg on the 18th. It was timed to coincide with the release of much of Llama-3, very much the approach of telling your story directly. Dwarkesh is now the true tech media. A meteoric rise, and well earned.

This is two related posts in one. First I cover the podcast, then I cover Llama-3 itself.

My notes are edited to incorporate context from later explorations of Llama-3, as I judged that the readability benefits exceeded the purity costs.

Podcast Notes: Llama-3 Capabilities

---

Outline:

(00:51) Podcast Notes: Llama-3 Capabilities

(03:09) The Need for Inference

(07:08) Great Expectations

(11:29) Open Source and Existential and Other Risks

(30:50) Interview Overview

(33:22) A Few Reactions

(47:53) Safety First

(54:15) Core Capability Claims

(56:11) How Good are the 8B and 70B Models in Practice?

(01:02:31) Architecture and Data

(01:05:08) Training Day

(01:09:17) What Happens Next With Meta's Products?

(01:12:24) What Happens Next With AI Thanks To These Two Models?

(01:14:04) The Bigger One: It's Coming

(01:14:59) Who Wins?

(01:17:21) Who Loses?

(01:21:49) How Unsafe Will It Be to Release Llama-3 400B?

(01:24:12) The Efficient Market Hypothesis is False

(01:27:09) What Next?

---

First published:

Source:

Narrated by TYPE III AUDIO.

I

Imagine an alternate version of the Effective Altruism movement, whose early influences came from socialist intellectual communities such as the Fabian Society, as opposed to the rationalist diaspora. Let's name this hypothetical movement the Effective Samaritans.

Like the EA movement of today, they believe in doing as much good as possible, whatever this means. They began by evaluating existing charities, reading every RCT to find the very best ways of helping.

But many effective samaritans were starting to wonder. Is this randomista approach really the most prudent? After all, Scandinavia didn’t become wealthy and equitable through marginal charity. Societal transformation comes from uprooting oppressive power structures.

The Scandinavian societal model which lifted the working class, brought weekends, universal suffrage, maternity leave, education, and universal healthcare can be traced back all the way to 1870's where the union and social democratic movements got their start.

In many developing countries [...]

---

Outline:

(00:03) I

(05:39) II

(10:19) III

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

---

First published:

Source:

Narrated by TYPE III AUDIO.

I examine the literature on transfer learning in humans. Far

When learning, one would like to progress faster, and learn things

I've spent ~20 hours reading and skimming papers

I've looked at [...]

---

Outline:

(03:49) Recommendations

(05:30) What I Am Looking For

(07:00) Candidate Interventions

(07:04) Effective Learning Techniques

(09:12) What Didnt Work

(09:59) Far Transfer

(12:44) Negative Transfer

(13:05) Error Management Training

(14:45) Pólya Method

(16:48) Accelerated Expertise

(18:56) Inducing Transfer

(19:43) Dual N-Back and Brain Training

(20:07) Judgmental Forecasting

(21:37) Creating Self-Improving Institutions

(24:23) How Id Proceed If I Was Further Researching Transfer

(25:00) Questions

(26:30) Context

(26:42) Appendix A: My Impression of the Literature

(27:25) See Also

The original text contained 6 footnotes which were omitted from this narration.

---

First published:

Source:

Narrated by TYPE III AUDIO.

1. If you find that you’re reluctant to permanently give up on to-do list items, “deprioritize” them instead

I hate the idea of deciding that something on my to-do list isn’t that important, and then deleting it off my to-do list without actually doing it. Because once it's off my to-do list, then quite possibly I’ll never think about it again. And what if it's actually worth doing? Or what if my priorities will change such that it will be worth doing at some point in the future? Gahh!

On the other hand, if I never delete anything off my to-do list, it will grow to infinity.

The solution I’ve settled on is a priority-categorized to-do list, using a kanban-style online tool (e.g. Trello). The left couple columns (“lists”) are very active—i.e., to-do list items that I might plausibly do today or tomorrow, with different columns for different contexts [...]

---

Outline:

(00:04) 1. If you find that you’re reluctant to permanently give up on to-do list items, “deprioritize” them instead

(02:12) 2. If you find that you’re reluctant to delete (or heavily edit) a piece of text / slide that you worked hard on, copy it into a “graveyard” first

(03:23) 3. If you find that you’re reluctant to throw out papers, make it fast and easy to file them

(05:16) 4. If you find that you’re reluctant to delete computer files / emails, don’t empty the trash

---

First published:

Source:

Narrated by TYPE III AUDIO.

tl;dr: Recently reported GPT-J experiments [1 2 3 4] prompting for definitions of points in the so-called "semantic void" (token-free regions of embedding space) were extended to fifteen other open source base models from four families, producing many of the same bafflingly specific outputs. This points to an entirely unexpected kind of LLM universality (for which no explanation is offered, although a few highly speculative ideas are riffed upon).

Work supported by the Long Term Future Fund. Thanks to quila for suggesting the use of "empty string definition" prompts, and to janus for technical assistance.

"Mapping the semantic void: Strange goings-on in GPT embedding spaces" presented a selection of recurrent themes (e.g., non-Mormons, the British Royal family, small round things, holes) in outputs produced by prompting GPT-J to define points in embedding space randomly sampled at various distances from the token embedding centroid. This was tentatively framed [...]

---

Outline:

(04:05) Models tested

(04:49) Key results

(05:14) 1. group (non-)membership

(06:41) selected examples

(08:51) centroid and empty string definitions

(11:17) specific groups encountered

(14:09) current thinking

(16:38) 2. Mormons

(18:08) GPT-3 and GPT-4 base outputs

(19:30) current thinking

(20:00) 3. Church of England

(22:28) 4. (non-)members of royal families

(22:33) Examples

(23:45) Current thinking

(24:32) 5. (non-)members of the clergy

(26:11) empty string definitions

(27:59) 6. holes in things

(28:04) Examples

(31:04) 7. small round things

(31:09) Examples

(33:52) current thinking/feeling

(35:16) 8. pieces of wood or metal

(35:20) Examples

(37:17) centroid definition

(37:51) Current thinking

(38:30) 9. (small) pieces of cloth

(38:35) Examples

(40:11) Current thinking

(40:39) 10. communists (and other political parties)

(40:44) Examples

(42:49) centroid definitions

(43:07) current thinking

(44:39) Examples

(45:48) GPT-3 glitch token outputs

(46:20) GPT-4 definitions

(46:58) Current thinking

(47:20) 12. being in a state of being

(47:25) Examples

(49:32) 13. an X that isn’t an X

(49:36) examples

(51:01) centroid definitions

(51:25) current thinking

(52:04) 14. the most important

(52:09) Examples

(53:31) 15. narrow geological features

(53:36) Examples

(54:39) 16. Small pieces of land

(57:09) Appendix A: complete results

(58:03) Appendix B: miscellaneous memorable outputs

(58:15) GPT-2-small

(58:44) GPT-2-xl

(59:04) Pythia-70m

(59:19) Pythia-160m

(59:47) Pythia-410m

(01:00:04) Pythia-1b

(01:00:25) Pythia-2.8b

(01:00:36) Pythia-6.9b

(01:00:46) Mistral-7b

(01:00:54) OpenLLaMa-3b

(01:01:15) OpenLLaMa-3b-v2

(01:01:23) OpenLLaMa-7b-v2

(01:01:35) StableLM-3b

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

---

First published:

Source:

Narrated by TYPE III AUDIO.