PODCAST

LessWrong (30+ Karma)

LessWrong

Audio narrations of LessWrong posts.

technology philosophy

All Episodes

8:30

“LLM in-context learning as (approximating)...

LessWrong (30+ Karma) ·

2025/06/06

Epistemic status: One week empirical project from a theoretical computer scientist. My analysis and presentation were both a little rushed; some information that would be interesting is missing from plots because I simply did not have time to include it. All known "breaking" issues are discussed and should not effect the conclusions. I may refine this post in the future. [This work was performed as my final project for ARENA 5.0.] Background I have seen several claims[1] in the literature that base

en-gb

3:23

[Linkpost] “Histograms are to CDFs as calibration...

LessWrong (30+ Karma) ·

2025/06/06

This is a link post. As you know, histograms are decent visualizations for PDFs with lots of samples...10k predictions, 20 bins ...but if there are only a few samples, the histogram-binning choices can matter a lot:10 predictions, 4 binssame 10 predictions, 7 bins The binning (a) discards information, and worse, (b) is mathematically un-aesthetic. But a CDF doesn't have this problem!same 10 predictions, every data point precisely represented If you make a bunch of predictions, and you want to know

en-gb

4:43

[Linkpost] “Discontinuous Linear Functions?!” by...

LessWrong (30+ Karma) ·

2025/06/06

This is a link post. We know what linear functions are. A function f is linear iff it satisfies additivity _f(x + y) = f(x) + f(y)_ and homogeneity _f(ax) = af(x)_.We know what continuity is. A function f is continuous iff for all ε there exists a δ such that if _|x - x_0|_ < δ, then _|f(x) - f(x_0)|_ < ε. An equivalent way to think about continuity is the sequence criterion: f is continuous

en-gb

16:29

“The Stereotype of the Stereotype” by Ike

LessWrong (30+ Karma) ·

2025/06/06

Epistemic Status: Attempt to codify new term. I want to define a new term that will be useful in the discourse. Imagine you landed in a foreign country with dogs, but with no word for dogs. You would see these things, and want to talk about them. You would like to be able to tell people whether they can bring dogs into your house, but you can’t, because there isn’t any word for “dog.” You’d like to ask if you can pet someone's dog, but you can’t do that either, because when you ask “can I pet

en-gb

9:32

“‘Flaky breakthroughs’ pervade coaching — and no one...

LessWrong (30+ Karma) ·

2025/06/04

Has someone you know ever had a “breakthrough” from coaching, meditation, or psychedelics — only to later have it fade? Show tweet For example, many people experience ego deaths that can last days or sometimes months. But as it turns out, having a sense of self can serve important functions (try navigating a world that expects you to have opinions, goals, and boundaries when you genuinely feel you have none) and finding a better cognitive strategy without downsides is non-trivial. Because the

en-gb

6:43

“Individual AI representatives don’t solve Gradual...

LessWrong (30+ Karma) ·

2025/06/04

Imagine each of us has an AI representative, aligned to us, personally. Is gradual disempowerment solved?[1] In my view, no; at the same time having AI representatives helps at the margin. I have two deep reasons for skepticism.[2] Here is the first one. Humans are Not Alone We, as individuals, are not the only agents or “agencies” in this world. Other goal-oriented strategic entities include states, corporations, and to some extent egregores. For the sake of argument, imagine the current

en-gb

19:17

“Broad-Spectrum Cancer Treatments” by sarahconstantin

LessWrong (30+ Karma) ·

2025/06/04

Midjourney, “engraving of Apollo shooting his bow at a distant cancer cell” Introduction and Principles The conventional wisdom is that we can’t “cure cancer” because every cancer is different. And it's true that cancer is genetically and biochemically diverse, and that targeted cancer therapies generally only work on very specific sub-populations. But some of the oldest, most reliable, and still most commonly used classes of cancer treatment -- radiotherapy and cytotoxic chemotherapy -- are also

en-gb

0:50

[Linkpost] “Seeing how well an agentic AI coding tool...

LessWrong (30+ Karma) ·

2025/06/03

This is a link post. When it comes to disputes about AI coding capabilities, I think a lot of people who either don't follow AI capabilities developments or who aren't particularly capable coders might be missing some context, so I wrote a post wherein I try to get an AI coding tool to reimplement a feature in my game then closely study the output to demonstrate how well it's able to do that. --- First published: June 1st, 2025 Source:

en-gb

25:51

“1. The challenge of unawareness for impartial...

LessWrong (30+ Karma) ·

2025/06/02

(This sequence assumes basic familiarity with longtermist cause prioritization concepts, though the issues I raise also apply to non-longtermist interventions.) Are EA interventions net-positive from an impartial perspective — one that gives moral weight to all consequences, no matter how distant? What if they’re neither positive, nor negative, nor neutral? Trying to reduce x-risk, improve institutions, or end factory farming might seem robustly positive. After all, we don’t need to be certain in

en-gb

7:48

“Unfaithful Reasoning Can Fool Chain-of-Thought...

LessWrong (30+ Karma) ·

2025/06/02

This research was completed for LASR Labs 2025 by Benjamin Arnav, Pablo Bernabeu-Pérez, Nathan Helm-Burger, Tim Kostolansky and Hannes Whittingham. The team was supervised by Mary Phuong. Find out more about the program and express interest in upcoming iterations here. Read the full paper: "CoT Red-Handed: Stress Testing Chain-of-Thought Monitoring." Chain-of-thought (CoT) monitoring—where safety systems review a model's intermediate reasoning steps—is gaining traction at frontier labs like Google

en-gb

23:20

“The Value Proposition of Romantic Relationships” by...

LessWrong (30+ Karma) ·

2025/06/02

What's the main value proposition of romantic relationships? Now, look, I know that when people drop that kind of question, they’re often about to present a hyper-cynical answer which totally ignores the main thing which is great and beautiful about relationships. And then they’re going to say something about how relationships are overrated or some such, making you as a reader just feel sad and/or enraged. That's not what this post is about. So let me start with some more constructive motivations…

en-gb

10:42

“The 80/20 playbook for mitigating AI scheming risks...

LessWrong (30+ Karma) ·

2025/06/01

Adapted from this twitter thread. See this as a quick take. Mitigation Strategies How to mitigate Scheming? Architectural choices: ex-ante mitigation Control systems: post-hoc containment White box techniques: post-hoc detection Black box techniques Avoiding sandbagging We can combine all of those mitigation via defense-in-depth system (like the Swiss Cheese model below) I think that applying all of those strategies should divide the risk by at least 3. 1. Architectural Choices: Ex-ante

en-gb

9:50

“The best approaches for mitigating ‘the intelligence...

LessWrong (30+ Karma) ·

2025/05/31

There have recently been various proposals for mitigations to "the intelligence curse" or "gradual disempowerment"—concerns that most humans would end up disempowered (or even dying) because their labor is no longer valuable. I'm currently skeptical that the typically highlighted prioritization and interventions are best and I have some alternative proposals for relatively targeted/differential interventions which I think would be more leveraged (as in, the payoff is higher relative to the

en-gb

12:29

“‘GiveWell for AI Safety’: Lessons learned in a week”...

LessWrong (30+ Karma) ·

2025/05/31

On prioritizing orgs by theory of change, identifying effective giving opportunities, and how Manifund can help. Epistemic status: I spent ~20h thinking about this. If I were to spend 100+ h thinking about this, I expect I’d write quite different things. I was surprised to find early GiveWell ‘learned in public’: perhaps this is worth trying. The premise: EA was founded on cost-effectiveness analysis—why not try this for AI safety, aside from all the obvious reasons¹? A good thing about early

en-gb

27:15

“Orphaned Policies (Post 5 of 6 on AI Governance)” by...

LessWrong (30+ Karma) ·

2025/05/30

In previous posts in this sequence, I laid out a case for why most AI governance research is too academic and too abstract to have much influence over the future. Politics is noisy and contested, so we can’t expect that good AI governance ideas will spread on their own – we need a large team of people who are actively promoting those ideas. Unfortunately, we currently have at least 3 researchers for every advocate, so many policy ideas have been “orphaned,” i.e., nobody is taking those ideas and

en-gb

3:49

“Do you even have a system prompt? (PSA)” by...

LessWrong (30+ Karma) ·

2025/05/30

Everyone around me has a notable lack of system prompt. And when they do have a system prompt, it's either the eigenprompt or some half-assed 3-paragraph attempt at telling the AI to “include less bullshit”. I see no systematic attempts at making a good one anywhere.[1] (For clarity, a system prompt is a bit of text—that's a subcategory of "preset" or "context"—that's included in every single message you send the AI.) No one says “I have a conversation with Claude, then edit the system prompt based

en-gb

2:45

[Linkpost] “Incorrect Baseline Evaluations Call into...

LessWrong (30+ Karma) ·

2025/05/30

This is a link post. There has been a flurry of recent papers proposing new RL methods that claim to improve the “reasoning abilities” in language models. The most recent ones, which show improvements with random or no external rewards have led to surprise, excitement and confusion. We analyzed 7 popular LLM RL papers (100+ to 3000+ likes, 50k+ to 500k+ views on X) including “Spurious Rewards”, “RL from 1 example”, and 3 papers exploring “Intrinsic Confidence Rewards”. We found that in most of these

en-gb

4:40

“CFAR is running an experimental mini-workshop (June...

LessWrong (30+ Karma) ·

2025/05/30

Hello from the Center for Applied Rationality! Some of you may have attended our classic applied rationality workshops in the past; others of you may have wanted to attend a workshop but not yet had a chance to. It's been a while since we've last run public-facing workshops, but we wanted to say: We're not dead! (More on that perhaps in a later post) We have a new experimental mini-workshop coming up soon and hopefully more workshop content to follow after! Our new workshop will be held after

en-gb

20:00

“Gradual Disempowerment: Concrete Research Projects”...

LessWrong (30+ Karma) ·

2025/05/29

This post benefitted greatly from comments, suggestions, and ongoing discussions with David Duvenaud, David Krueger, and Jan Kulveit. All errors are my own. A few months ago, I and my coauthors published Gradual Disempowerment (GD hereafter). It was mostly about how things might go wrong, but naturally a lot of the resulting interest has been about solutions. We have some more formal followup work coming: in the meantime, this is my 80/20 for ‘what would I do if I had way more time’ / ‘what would

en-gb

123:22

“Truth or Dare” by Duncan Sabien (Inactive)

LessWrong (30+ Karma) ·

2025/05/29

Author's note: This is my apparently-annual "I'll put a post on LessWrong in honor of LessOnline" post. These days, my writing goes on my Substack. There have in fact been some pretty cool essays since last year's LO post. Structural note: Some essays are like a five-minute morning news spot. Other essays are more like a 90-minute lecture. This is one of the latter. It's not necessarily complex or difficult; it could be a 90-minute lecture to seventh graders (especially ones with the right

en-gb

4,489 results

LessWrong (30+ Karma)

All Episodes

Similar Podcasts

Talking Tech w/ Wayne

Wayne Hunter

The Everything Feed - All Packet Pushers Pods

Packet Pushers

The Good Tech Companies

HackerNoon

Engadget News + Next

Engadget

Cybersecurity Tech Brief By HackerNoon

HackerNoon