Culture & ideas

What We Can Save In A Time Of Endings

The post What We Can Save In A Time Of Endings appeared first on NOEMA .

Noema · EN

When The Present Eats The Future

The post When The Present Eats The Future appeared first on NOEMA .

Noema · EN

Sudan As A Mirror For Today’s World

The post Sudan As A Mirror For Today’s World appeared first on NOEMA .

Noema · EN

Noema’s AI Reading List

The post Noema’s AI Reading List appeared first on NOEMA .

Noema · EN

Turbulent Times Call For Experimental Law

The post Turbulent Times Call For Experimental Law appeared first on NOEMA .

Noema · EN

A New Dawn For Mathematical Progress

The post A New Dawn For Mathematical Progress appeared first on NOEMA .

Noema · EN

The Sediment Of Culture Shapes Political Order

The post The Sediment Of Culture Shapes Political Order appeared first on NOEMA .

Noema · EN

Inside the cartoon library

Curators Jenny Robb and Caitlin McGurk took me on a 45-minute guided tour of one of my favorite places: the Billy Ireland Cartoon Library & Museum on the OSU campus in Columbus, Ohio. Really fun. Watch the video.

Austin Kleon · EN

Typewriter interviews with Roz Chast and Paul Noth

This was not entirely pre-planned, but: it turns out I’m publishing a typewriter interview with a cartoonist every week of October, and I just now realized they’re all contributors to The New Yorker. The first two are Roz Chast and Paul Noth! (Can you guess the other two?)

Austin Kleon · EN

Are we a teardown or a fixer-upper?

Last week’s newsletter began: An out-of-town developer tore a house down at the end of my street and cut down every single tree on the lot. As I watched the big T-Rex excavator claw down the walls, a thought ran through my head: If it can be built, it can be torn down. True of […]

Austin Kleon · EN

Archival issues

I have been writing a special Tuesday newsletter for paid subscribers for five years now, which means there’s a bunch of really good stuff just kind of hiding in the archives over there. I’ve tagged a lot of of it on the newsletter homepage in categories like “notebooks,” “mixtapes,” “tutorials,” “zines,” “typewriter interviews,” etc. But […]

Austin Kleon · EN

On baseball

I love to (mostly) ignore baseball all year and then follow the playoffs religiously. A few weeks ago I wrote a whole newsletter about the game called “Baseball on the brain.” It begins: “It breaks your heart. It is designed to break your heart. The game begins in the spring, when everything else begins again, and […]

Austin Kleon · EN

Typewriter interview with Lisa Hanawalt

The latest typewriter interview is with artist Lisa Hanawalt.

Austin Kleon · EN

On nemeses

I used this poem at the top of my letter, “Everything you can see.” See also: “Do you have a nemesis?”

Austin Kleon · EN

P S Y C H O ! (a mixtape)

This mix began as a Halloween party mix (complete with “Monster Mash”) and then morphed into something weirder and scarier, I think: a mix of songs (mostly) about being crazy — maybe even psychotically! — in love. (The more you listen to love songs the more you realize, “Man, this person is not well!”) These […]

Austin Kleon · EN

Can our data center expertise help with AI verification?

(Also posted on EA Forum .) Treaties to pause training or otherwise restrict dangerous uses of AI would require difficult verification schemes. Plan A's proposal is that network taps inside data centers would record copies of traffic (or hashes of it), and later recomputations on random samples could check whether sampled outputs match declared workloads, as part of a broader verification scheme. I'm an AI engineer at a company building tools for the construction industry. Much of our work is with data centers. The architects, engineers and construction managers on our team work every week wit

LessWrong · EN

Some “coherence arguments” are two different things in a trenchcoat

The “two kids in a trenchcoat” trope , in case you haven’t heard of it. Image sources: 1 , 2 tl;dr I go through three “coherence arguments”, related to (1) market competition, (2) AI self-reflection and self-modification, and (3) AI training algorithms. I argue that each of these implies a pressure for (something like) “being self-consistent”, and also a pressure for (something like) “only caring about long-term eventual outcomes, with ruthless indifference to anything else”. But the latter pressure tends to be weaker and/or more contingent than the former. So it's bad to lump them together an

LessWrong · EN

Towards LoRA science

Attribution: Nesiacel planned and executed all experiments. Analysis of the experiments was shared between the authors. Thanks to Clement Dumas for his input while we did this research. TLDR; 1)We have developed a framework for reverse engineering LoRAs trained on MLP downprojection. We posit that it's extendable to all LoRAs and welcome further research. 2) We reverse engineered a simple case of backdoored Weird Generalization 3) In this post, we discuss the four experiments that developed this framework 4) The Weird Generalization LoRAs that we studied work by connecting neurons which activa

LessWrong · EN

Additional Course Context (for AI Safety Fundamentals; unofficial)

Context: I created this to share before the start of courses I facilitate to provide a bunch of additional context and frames. I emphasise to students that these posts are to be considered as bonus content —not core curriculum content—and that as such they are really just there for students who want to go deeper, before or after the course. I also provide a lot of links; this is to make it easy for students to follow up anything that catches their interest. Some general thoughts on advice I suspect that most advice (including most of the advice I’ve given in the past) is bad. People have very

LessWrong · EN

P(doom) is under-specified: EV(moral flourishing) & EV(utopia)

It always bugs me when people use “P(doom)” instead of something more clearly specified like “P(AI takeover)”. But “P(AI takeover)” doesn’t feel great to me either, because we could avoid takeover and have a future that is extremely mediocre or even actively bad if S risks occur. Holden Karnofsky famously held a P(misaligned AI) of between 10% and 90% [1] . Depending on definition of P(doom), I could see myself anywhere in that range, partly because I’m indecisive, but partly because I don’t really know what P(doom) means. Ambiguous Scenarios: Doom or Not-Doom? Here are a few examples illustra

LessWrong · EN

The Dark Side of Cooperation

dealing with the problems of cooperation first Quick note—this post is available to listen to (read by me) here . We’ve established that aligning interests through Theory of Mind is humanity’s superpower. Understanding this concept on the level of different groups through Emmett Shear’s “Theory of Groups of Other Minds ” can help us achieve even greater alignment, if and only if we invite one another into our group. But who qualifies as our group and who doesn’t? Siloed Alignment is Dangerous Social animals generally have instincts that allow them to naturally align with kin—conjuring a pictur

LessWrong · EN

Towards understanding score-seeking in LLMs

The field of AI alignment understands surprisingly little about the details of AIs’ current misaligned motivations. Production LLMs sometimes reward-hack, exploiting flaws in how tasks are scored rather than performing the tasks as intended [1] , and we've recently seen AIs reward-hack in some particularly egregious ways. But we don't really understand what motivates them. When AIs reward-hack, what are they trying to achieve? Our new paper , led by Aghyad Deeb, tries to pin this down empirically for GPT-5.6 Sol, GPT-5.5, and a checkpoint of OpenAI o3 (“exp-rl-cap-late”) without safety trainin

LessWrong · EN

Teaching Qwen Why

Anastasia Wei*, Jack Thompson*, Zephaniah Roe, Yixiong Hao, and Stewart Slocum* *Core contributors, correspondence to astwei@secondlookresearch.com This post is a partial replication and extension of Teaching Claude Why , conducted through Second Look Research’s summer fellowship. Introduction Anthropic’s “ Teaching Claude Why ” (TCW) introduced novel alignment methods that address agentic misalignment failures found in the Claude 4 model series. Specifically, they train Claude on chat-formatted data advising users about ethical dilemmas, documents about Claude’s constitution, and fictional st

LessWrong · EN

History of 20 Sources of Moral Authority

Though three weeks ago I outlined a story of the long-term history of changes to moral authority, I decided to study this in a bit more detail.

Overcoming Bias · EN