Google Deepmind has announced Gemini Diffusion. Though buried under a host of other IO announcements it's possible that this is actually the most important one!

This is significant because diffusion models are entirely different to LLMs. Instead of predicting the next token, they iteratively denoise all the output tokens until it produces a coherent result. This is similar to how image diffusion models work.

I've tried they results and they are surprisingly good! It's incredibly fast, averaging nearly 1000 tokens a second. And it one shotted my Google interview question, giving a perfect response in 2 seconds (though it struggled a bit on the followups).

It's nowhere near as good as Gemini 2.5 pro, but it knocks ChatGPT 3 out the water. If we'd seen this 3 years ago we'd have been mind blown.

Now this is wild for two reasons:

1. We now have [...]

---

First published:

Source:

Linkpost URL:
https://deepmind.google/models/gemini-diffusion/

---

Narrated by TYPE III AUDIO.

Have the Accelerationists won?

Last November Kevin Roose announced that those in favor of going fast on AI had now won against those favoring caution, with the reinstatement of Sam Altman at OpenAI. Let's ignore whether Kevin's was a good description of the world, and deal with a more basic question: if it were so—i.e. if Team Acceleration would control the acceleration from here on out—what kind of win was it they won?

It seems to me that they would have probably won in the same sense that your dog has won if she escapes onto the road. She won the power contest with you and is probably feeling good at this moment, but if she does actually like being alive, and just has different ideas about how safe the road is, or wasn’t focused on anything so abstract as that, then whether she ultimately wins or [...]

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

Summary

Porn content has gotten more extreme over time. Here's the average title for the first full year of Pornhub's existence, 2008:

• "Hot blonde girl gets fucked"

and here's the average title for 2023:

• "FAMILYXXX - "I Cant Resist My Stepsis Big Juicy Ass" (Mila Monet)"

Why did this change happen? We can understand porn's progression by converting titles to language embeddings. I downloaded Internet Archive snapshots of "pornhub.com" from 2008 - 2023 and analyzed the embeddings of the titles on the main page.

I found three distinct eras of titling: 2008-2009, 2010-2016, 2017-present. The current trend, since 2017, is characterized mainly by an emphasis on incest and other sexual violence.

Titles are generally representative of actual video content, and provide a reasonable heuristic for measuring actual content change, though some SEO effects exist.

The conclusion is a slightly ominous one: we are close to semantic bedrock [...]

---

Outline:

(00:12) Summary

(01:32) Data and Methods

(03:17) Title Accuracy

(04:32) Calculating Yearly Centroids

(04:57) Centroid Similarity

(05:32) Centroid Clusters

(06:22) Centroid Titles

(08:57) Keyword Trends

(11:30) t-SNE Clusters

(11:55) Haircolor

(12:22) Pornstar Names

(12:50) Violence

(13:14) Women

(13:36) Men

(14:02) Racial

(14:25) Manufacturing

(15:02) Benign

(15:29) Sexual Violence

(16:58) Conclusions

---

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.

Our government, having withdrawn the new diffusion rules, has now announced an agreement to sell massive numbers of highly advanced AI chips to UAE and Saudi Arabia (KSA). This post analyzes that deal and that decision.

It is possible, given sufficiently strong agreement details (which are not yet public and may not be finalized) and private unvoiced considerations, that this deal contains sufficient safeguards and justifications that, absent ability to fix other American policy failures, this decision is superior to the available alternatives. Perhaps these are good deals, with sufficiently strong security arrangements that will actually stick.

Perhaps UAE and KSA are more important markets and general partners than we realize, and the rest of the world really is unable to deploy capital and electrical power the way they can and there is nothing we can do to change this, and perhaps they have other points of [...]

---

Outline:

(02:28) Choosing Sides In the War on Cancer

(03:23) The Central Points From Last Week

(07:34) Diffusion Controls Have Proven Vital

(10:27) It's a Huge Deal

(14:11) Do You Feel Secure?

(17:09) Why just count server racks fails

(18:29) Bottom-line probability estimate

(21:41) Semianalysis Defends the Deal

(28:09) Understanding the China Hawks

(35:35) Rhetoric Unbecoming

(37:18) Could China Have 'Done This Deal'?

(38:37) Tyler Cowen Asks Good Questions

(43:38) Saudi Arabia Also Made a Deal

(45:35) At Best A Second Best Solution

---

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.

On January 28, 2025 (during the pontificate of Pope Francis) the Catholic Church put out its position paper on AI: Antiqua et nova: Note on the Relationship Between Artificial Intelligence and Human Intelligence. Here are my thoughts, which I wrote back in February, but not much has changed since then, and the new pope has continued citing AI as of particular concern to the Catholic Church.

"With all your modern science, are you any closer to understanding the mystery of how a robot walks or talks?" "Yes, you idiot. The circuit diagram is right here on the inside of your case!" (Futurama)

Four characteristics of humanity

Section III is the real meat of the document. It lists four essential characteristics, and argues that AI falls short of the fullness of humanity in each respect. These are: rationality, truth-seeking, embodiment, and relationality.

I will address embodiment [...]

---

Outline:

(00:54) Four characteristics of humanity

(01:33) Embodiment and relationality

(11:59) Rationality

(14:31) Truth-seeking

(15:59) How much of Antiqua et nova is specifically theistic, or Christian?

(16:52) What kind of theism is the document based on?

(18:16) Besides that, what specifically Christian points are made?

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

(h/t Otis Reid)

I think this post captures a lot of important features of the US policymaking system. Pulling out a few especially relevant/broadly applicable sections:

1. There's No Efficient Market For Policy

There can be a huge problem that nobody is working on; that is not evidence that it's not a huge problem. Conversely, there can be a marginal problem swamped with policy work; that's not evidence it's really all that big of a deal.

On the upside, this means there are never-ending arbitrage opportunities in policy. Pick your workstreams wisely.

2. Personnel Really Is The Most Important Thing

The quality of staffers varies dramatically and can make or break policy efforts. Some Hill staffers are just awesome; if they like your idea, they'll take it and run with it, try to find the right cosponsors, understand where it fits procedurally, etc. Other staffers [...]

---

First published:

Source:

Linkpost URL:
https://www.greentape.pub/p/one-year-in-dc

---

Narrated by TYPE III AUDIO.

The Future of Life Foundation is launching a fellowship on AI for Human Reasoning.

Fellowship on AI for Human Reasoning

Apply by June 9th | $25k–$50k stipend | 12 weeks, from July 14 - October 3

Join us in working out how to build a future which robustly empowers humans and improves decision-making.

FLF's incubator fellowship on AI for human reasoning will help talented researchers and builders start working on AI tools for coordination and epistemics. Participants will scope out and work on pilot projects in this area, with discussion and guidance from experts working in related fields. FLF will provide fellows with a $25k–$50k stipend, the opportunity to work in a shared office in the SF Bay Area, and other support.

In some cases we would be excited to provide support beyond the end of the fellowship period, or [...]

---

First published:

Source:

Linkpost URL:
https://www.flf.org/fellowship

---

Narrated by TYPE III AUDIO.

Confidence notes: I am a physicist working on computational material science, so I have some familiarity with the field, but don't know much about R&D firms or economics. Some of the links in this article were gathered from a post at pivot-to-ai.com and the BS detector.

The paper "Artificial Intelligence, Scientific Discovery, and Product Innovation" was published as an Arxiv preprint last December, roughly 5 months ago, and was submitted to a top economics journal.

The paper claimed to show the effect of an experiment at a large R&D company. It claimed the productivity of a thousand material scientists was tracked before and after the introduction of an machine learning material generation tool. The headline results was that the AI caused a 44% increase in materials discovery at the firm, with a productivity increase of 81% for top-decile scientists.

This research was breathlessly reported on in [...]

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

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

CGP Grey describes a phenomenon I think of as 'dreams of ideas' that I find useful as a tool to know when to stop working on a project.

You can be working on something and you are thinking about what it could be, but what is hard to know is if that thought in your head is more like when your brain dreams.

There is a phenomenon that I always think of as like, you wake up and you think you had a great idea in your dream, but what you really had was the idea of a great idea. You didn't have a great idea, but you had all the emotional resonance to it being a good idea in your dream. You're like, "Oh, why can't I think of the details?". It's like, because there weren't any details there at all...

Sometimes there just isn't the [...]

---

Outline:

(01:45) When to cull ideas

(02:19) Examples

(04:26) Not every unfulfilled idea is a dream

(05:06) The solution

(05:49) TL;DR

---

First published:

Source:

---

Narrated by TYPE III AUDIO.

---

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.