LessWrong posts by zvi

By zviSociety & Culture Technology Philosophy

LessWrong posts by zvi episodes

July 30, 2025“Childhood and Education: College Admissions” by Zvi
Table of Contents
College Applications.
The College Application Essay (Is) From Hell.
Don’t Guess The Teacher's Password, Ask For It Explicitly.
A Dime a Dozen.
Treat Admissions Essays Like Games of Balderdash.
It's About To Get Worse.
Alternative Systems Need Good Design.
The SAT Scale Is Broken On Purpose.
College Applications
In case you missed it, yes, of course Harvard admissions are way up and Harvard did this on purpose. The new finding is that Harvard was recruiting African American applicants in particular, partly in order to balance conditional acceptance rates. One could of course also argue that the goal was ‘to find more worthy students,’ with the counterevidence being that the test scores of such applicants declined as more applications came in (as they obviously would for any group) but the scores of those who got [...]

---
Outline:
(00:11) College Applications
(00:57) The College Application Essay (Is) From Hell
(04:35) Don't Guess The Teacher's Password, Ask For It Explicitly
(09:11) A Dime a Dozen
(10:18) Treat Admissions Essays Like Games of Balderdash
(14:54) It's About To Get Worse
(23:20) Alternative Systems Need Good Design
(29:07) The SAT Scale Is Broken On Purpose

---

First published:
July 30th, 2025

Source:
https://www.lesswrong.com/posts/yjXtoq86sz5LaRxDZ/childhood-and-education-college-admissions

---

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.
34 min
July 29, 2025 “Spilling the Tea” by Zvi
The Tea app is or at least was on fire, rapidly gaining lots of users. This opens up two discussions, one on the game theory and dynamics of Tea, one on its abysmal security.
It's a little too on the nose that a hot new app that purports to exist so that women can anonymously seek out and spill the tea on men, which then puts user information into an unprotected dropbox thus spilling the tea on the identities of many of its users.
In the circles I follow this predictably led to discussions about how badly the app was coded and incorrect speculation that this was linked to vibe coding, whereas the dumb mistakes involved were in this case fully human.
There was also some discussion of the game theory of Tea, which I found considerably more interesting and fun, and which will take up [...]

---
Outline:
(00:54) Pouring Some Tea
(03:12) Zero Tolerance Policy
(04:26) Solve For The Equilibrium
(06:15) A Fun Tale
(07:05) Is That Bad?
(15:57) Levels of Friction
(16:48) Can Hack It
(21:19) Should You Spill The Tea?

---

First published:
July 29th, 2025

Source:
https://www.lesswrong.com/posts/HNtXhv9yvqBYaSbgt/spilling-the-tea

---

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.
23 min
July 28, 2025“AI Companion Piece” by Zvi
AI companions, other forms of personalized AI content and persuasion and related issues continue to be a hot topic. What do people use companions for? Are we headed for a goonpocalypse? Mostly no, companions are used mostly not used for romantic relationships or erotica, although perhaps that could change. How worried should we be about personalization maximized for persuasion or engagement?
Table of Contents
Persuasion Should Be In Your Preparedness Framework.
Personalization By Default Gets Used To Maximize Engagement.
Companion.
Goonpocalypse Now.
Deepfaketown and Botpocalypse Soon.
Persuasion Should Be In Your Preparedness Framework
Kobi Hackenburg leads on the latest paper on AI persuasion.
Kobi Hackenberg: RESULTS (pp = percentage points):
1⃣Scale increases persuasion, +1.6pp per OOM
2⃣Post-training more so, +3.5pp
3⃣Personalization less so, <1pp
4⃣Information density drives persuasion gains
5⃣Increasing persuasion decreased factual accuracy
6⃣Convo > static, +40%
Zero is on [...]
---
Outline:
(00:35) Persuasion Should Be In Your Preparedness Framework
(04:33) Personalization By Default Gets Used To Maximize Engagement
(13:04) Companion
(16:50) Goonpocalypse Now
(24:09) Deepfaketown and Botpocalypse Soon

---

First published:
July 28th, 2025

Source:
https://www.lesswrong.com/posts/Rm3FD6646NDSCGxqM/ai-companion-piece

---

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.
29 min
July 25, 2025 “America’s AI Action Plan Is Pretty Good” by Zvi
No, seriously. If you look at the substance, it's pretty good.
I’ll go over the whole thing in detail, including the three executive actions implementing some of the provisions. Then as a postscript I’ll cover other reactions.
The White House Issues a Pretty Good AI Action Plan
There is a lot of the kind of rhetoric you would expect from a Trump White House. Where it does not bear directly on the actual contents and key concerns, I did my absolute best to ignore all the potshots. The focus should stay on the actual proposals.
The actual proposals, which are the part that matters, are far superior to the rhetoric.
This is a far better plan than I expected. There are a few points of definite concern, where the wording is ambiguous and one worries the implementation could go too far. Two in particular are the call for ensuring [...]
---
Outline:
(00:25) The White House Issues a Pretty Good AI Action Plan
(03:20) Pillar I: Accelerate AI Innovation
(03:25) Remove Red Tape and Onerous Regulations
(05:56) Ensure That Frontier AI Protects Free Speech And American Values
(12:36) Encourage Open-Source and Open-Weight AI
(14:17) Enable AI Adoption
(15:08) Empower American Workers In The Age of AI
(16:01) Support Next-Generation Manufacturing
(16:18) Invest in AI-Enabled Science and Build World-Class Scientific Datasets
(17:54) Invest in AI Interpretability, Control and Robustness Breakthroughs
(18:35) Build an AI Evaluations Ecosystem
(19:31) Accelerate AI Adoption in Government and in DOD
(21:32) Protect Commercial and Government AI Innovations
(21:54) Combat Synthetic Media in the Legal System
(22:24) Pillar II: Build American AI Infrastructure
(22:45) Create Streamlined Permitting for Data Centers, Semiconductor Manufacturing Facilities, and Energy Infrastructure while Guaranteeing Security
(23:59) Develop a Grid to Match the Pace of AI Innovation
(26:21) Restore American Semiconductor Manufacturing
(27:16) Train a Skilled Workforce for AI Infrastructure
(29:24) Bolster Critical Infrastructure Cybersecurity
(29:54) Promote Secure-By-Design AI Technologies and Applications
(31:04) Promote Mature Federal Capacity for AI Incident Response
(32:00) Pillar III: Lead in International AI Diplomacy and Security
(32:06) Export American AI to Allies and Partners
(34:24) Counter Chinese Influence in International Governance Bodies
(35:24) Strengthen AI Compute Export Control Enforcement
(36:00) Plug Loopholes in Existing Semiconductor Manufacturing Export Controls
(36:15) Align Protection Measures Globally
(37:21) Ensure that the U.S. Government is at the Forefront of Evaluating National Security Risks in Frontier Models
(38:36) Invest in Biosecurity
(39:10) Part 2: Reactions
(39:15) Speechless
(40:23) We Can All Just Get Along
(42:27) Okay Maybe Not All Of Us But What You Gonna Do
(45:27) Good Vibes Only
(49:00) Dare Not Speak Its Name
(50:09) Other Reactions In Brief
(51:51) Remarks of Note
(52:24) Conclusion: Helpful Is The Watchword

---

First published:
July 25th, 2025

Source:
https://www.lesswrong.com/posts/TTzyky2EQfYapxvMn/america-s-ai-action-plan-is-pretty-good

---

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.
55 min
July 24, 2025 “AI #126: Go Fund Yourself” by Zvi
The big AI news this week came on many fronts.
Google and OpenAI unexpectedly got 2025 IMO Gold using LLMs under test conditions, rather than a tool like AlphaProof. How they achieved this was a big deal in terms of expectations for future capabilities.
ChatGPT released GPT Agent, a substantial improvement on Operator that makes it viable on a broader range of tasks. For now I continue to struggle to find practical use cases where it is both worth using and a better tool than alternatives, but there is promise here.
Finally, the White House had a big day of AI announcements, laying out the AI Action Plan and three executive orders. I will cover that soon. The AI Action Plan's rhetoric is not great, and from early reports the rhetoric at the announcement event was similarly not great, with all forms of safety considered so [...]

---
Outline:
(01:41) Language Models Offer Mundane Utility
(03:04) Language Models Don't Offer Mundane Utility
(10:48) Huh, Upgrades
(11:40) 4o Is An Absurd Sycophant
(15:05) On Your Marks
(17:32) Choose Your Fighter
(18:12) When The Going Gets Crazy
(22:25) They Took Our Jobs
(26:36) Fun With Media Generation
(27:06) The Art of the Jailbreak
(30:15) Get Involved
(30:58) Introducing
(31:24) In Other AI News
(33:30) Show Me the Money
(40:39) Go Middle East Young Man
(48:21) Economic Growth
(49:40) Quiet Speculations
(57:44) Modest Proposals
(58:36) Predictions Are Hard Especially About The Future
(01:01:52) The Quest for Sane Regulations
(01:06:25) Chip City
(01:07:29) The Week in Audio
(01:07:50) Congressional Voices
(01:09:02) Rhetorical Innovation
(01:14:51) Grok Bottom
(01:18:22) No Grok No
(01:19:06) Aligning a Smarter Than Human Intelligence is Difficult
(01:22:36) Preserve Chain Of Thought Monitorability
(01:30:03) People Are Worried About AI Killing Everyone
(01:31:15) The Lighter Side

---

First published:
July 24th, 2025

Source:
https://www.lesswrong.com/posts/ygND532h4CotfPcp7/ai-126-go-fund-yourself

---

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.
1 hr 33 min
July 23, 2025 “GPT Agent Is Standing By” by Zvi
OpenAI now offers 400 shots of ‘agent mode’ per month to Pro subscribers.
This incorporates and builds upon OpenAI's Operator. Does that give us much progress? Can it do the thing on a level that makes it useful?
So far, it does seem like a substantial upgrade, but we still don’t see much to do with it.
What Is The Thing?
Greg Brockman (OpenAI): When we founded OpenAI (10 years ago!!), one of our goals was to create an agent that could use a computer the same way as a human — with keyboard, mouse, and screen pixels.
ChatGPT Agent is a big step towards that vision, and bringing its benefits to the world thoughtfully.
ChatGPT Agent: our first AI with access to a text browser, a visual browser, and a terminal.
Rolling out in ChatGPT Pro, Plus, and Team [July 17].
[...]
---
Outline:
(00:30) What Is The Thing?
(02:43) Can It Do The Thing?
(04:25) Access Granted
(06:29) Is The Thing Worth Doing?
(09:50) Danger, Will Robinson
(15:56) Some Potential Issues with MCP
(22:30) So Far a Whisper

---

First published:
July 23rd, 2025

Source:
https://www.lesswrong.com/posts/qC6owmPE3xpag3Wyi/gpt-agent-is-standing-by

---

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.
25 min
July 22, 2025 “Google and OpenAI Get 2025 IMO Gold” by Zvi
Congratulations, as always, to everyone who got to participate in the 2025 International Mathematical Olympiad, and especially to the gold and other medalists. Gautham Kamath highlights 11th grader Warren Bei, who in his 5th (!) IMO was one of five participants with a perfect 42/42 score, along with Ivan Chasovskikh, Satoshi Kano, Leyan Deng and Hengye Zhang.
Samuel Albanie: Massive respect to the students who solved P6.
Congratulations to Team USA, you did not ‘beat China’ but 2nd place is still awesome. Great job, China, you got us this time, three perfect scores is crazy.
You’ve all done a fantastic, amazingly hard thing, and as someone who tried hard to join you and only got as far as the [****, year censored because oh man I am old] USAMO and would probably have gotten 0/45 on this IMO if I had taken it today, and know [...]

---
Outline:
(01:25) This Is About AI And It Is Kind Of A Big Deal
(02:42) Skeptics Prematurely Declare Victory
(06:02) DeepMind Declares Victory
(10:10) OpenAI Declares Victory
(16:24) Bro I Don't Know
(17:19) Not So Fast?
(23:52) Not Announcing So Fast
(29:34) Are We Still Waiting On Anyone Else?
(30:41) This Was Not Widely Expected
(34:10) Jevons Paradox Strikes Again
(35:42) Nothing To Worry About
(45:52) So What If It Can Do The Math
(47:20) The Challenge Bet
(48:07) How \[I, Paul, would\] update
(49:39) Actually This Seems Like A Big Deal
(53:25) People On The Internet Sometimes Lie

---

First published:
July 22nd, 2025

Source:
https://www.lesswrong.com/posts/ZkgaPopsBkQgeA2k8/google-and-openai-get-2025-imo-gold

---

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.
59 min
July 21, 2025 “Monthly Roundup #32: July 2025” by Zvi
Welcome to the monthly roundup of things that don’t fit into other categories and don’t rise to the level of their own posts.
Bad News
When people tell you who they are, believe them (with obvious exceptions). In particular, if they explicitly describe themselves as evil, or demonic, or uses other similar terms, definitely believe them.
Did you know 67% of all college students bet on sports? That's a group that is majority female, so that statistic is wild. This is in the context of Ron Yorko developing a class in sports betting awareness from a neuroscience perspective for CMU freshman.
Cooking scales well, but for single people the economics are remarkably bad. Stop telling single people not to order delivery.
Chase Sapphire Reserve annual fee increases to $795 from $550, you get a $300 travel credit. That should cut down considerably the number [...]

---
Outline:
(00:18) Bad News
(05:12) Government Working
(07:51) Jones Act Watch
(11:23) Prize Fund
(13:19) Dining Out
(16:18) While I Cannot Condone This
(23:57) Good News, Everyone
(29:49) Opportunity Knocks
(30:00) Antisocial Media
(35:25) Technology Advances
(38:02) For Science!
(39:03) Various Things On Fire
(43:07) You Need Bigger Nametags
(44:19) Variously Effective Altruism
(49:04) For Your Entertainment
(58:46) Game Theory
(01:00:37) Gamers Gonna Game Game Game Game Game
(01:05:53) Who Wants To Be An American Citizen?
(01:07:15) I Was Promised Flying Self-Driving Cars
(01:09:53) Sports Go Sports
(01:10:42) The Lighter Side

---

First published:
July 21st, 2025

Source:
https://www.lesswrong.com/posts/zs4oDeNmKRukS7mjh/monthly-roundup-32-july-2025

---

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.
1 hr 15 min
July 18, 2025“On METR’s AI Coding RCT” by Zvi
METR ran a proper RCT experiment seeing how much access to Cursor (using Sonnet 3.7) would accelerate coders working on their own open source repos.
Everyone surveyed expected a substantial speedup. The developers thought they were being substantially sped up.
Instead, it turned out that using Cursor slowed them down.
That surprised everyone, raising the question of why.
Currently our best guess is this comes down to a combination of two factors:
Deeply understood open source repos are close to a worst-case scenario for AI tools, because they require bespoke outputs in various ways and the coder has lots of detailed local knowledge of the codebase that the AI lacks.
The coders in question mostly did not have experience with similar AI tools. The lack of a learning curve during the experiment challenges this, but the tools very clearly have a [...]

---
Outline:
(01:27) Epic Fail
(02:42) The Core Result
(07:10) Okay So That Happened
(12:21) Beginner Mindset
(19:43) Overall Takeaways

---

First published:
July 18th, 2025

Source:
https://www.lesswrong.com/posts/m2QeMwD7mGKH6vDe2/on-metr-s-ai-coding-rct

---

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.
22 min
July 17, 2025 “AI #125: Smooth Criminal” by Zvi
One story has towered over things this week. Unleash the Grok also known as the anime waifu codependent AI girlfriend Ani, also known as MechaHitler, or worse, they did. There's several sections here with more follow-ups. We also got the excellent model Kimi K2.
Perhaps quietly an even bigger story, and bigger fail, is the announced intention by the Trump administration to allow Nvidia to resume selling H20 AI chips to China. There may still be time to stop this, if not it is a very large unforced error, and it allows us to narrow down what it is our current administration primarily cares about, since we already know it isn’t ‘keep humans in control of the future and alive’ and this strongly suggests it is also not America or to ‘beat China.’
Another quiet but big development was the release of the surprisingly excellent new EU General-Purpose Code of [...]
---
Outline:
(01:16) Language Models Offer Mundane Utility
(03:24) Language Models Don't Offer Mundane Utility
(04:58) o3 Is a Lying Liar
(05:34) Thanks For The Memories
(06:23) Huh, Upgrades
(08:24) Choose Your Fighter
(08:39) Deepfaketown and Botpocalypse Soon
(16:12) They Took Our Jobs
(17:40) The Art of the Jailbreak
(18:55) Get Involved
(19:38) Introducing
(20:03) In Other AI News
(25:54) Bullshit On Bullshit
(29:04) Safety First
(31:27) Vitalik Praises And Critiques AI 2027
(39:58) Show Me the Money
(50:45) Quiet Speculations
(51:11) The Right Questions
(59:39) The Quest for Sane Regulations
(01:03:17) Chip City
(01:15:24) The Week in Audio
(01:16:14) Rhetorical Innovation
(01:20:29) Incorrectly Feeling The AGI
(01:21:19) RIP Absurdity Heuristic
(01:24:15) Worse Than MechaHitler
(01:30:05) An Alignment Problem
(01:36:59) Aligning a Smarter Than Human Intelligence is Difficult
(01:47:58) The Lighter Side

---

First published:
July 17th, 2025

Source:
https://www.lesswrong.com/posts/6Ykx8GpbDaBstjyBD/ai-125-smooth-criminal

---

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.
1 hr 51 min