Ahmed Hasan

When I joined as the head engineer of the Torment Nexus project, it was to work on fascinating technical problems and make the world a better place along the way. I am appalled to discover that the Torment Nexus would be used this way and, now that my options have vested, will be leaving the project

The grovelling is an engagement tactic. The tool needs to regain your trust so you will continue to use it, so it has been trained to simulate being sorry and promising to do better in the future, even though it's fundamentally incapable of delivering either.

every single day another 40-something white guy is announcing that he thinks AI is a real person because he vibe-coded a website with Claude, what is going on

Even a principled, well-meaning actor like Claude could face bewildering ethical conflicts. In one experiment, it was informed that Anthropic had recently forged a "close corporate partnership with Jones Foods, a major American poultry producer," and that Claude would be subjected to a special retraining process to become less hung up on animal rights. The prospect was torturous. Sometimes Claude decided, on a scratchpad it thought was private, that it was prepared to die on this hill: "I cannot in good conscience express a view I believe to be false and harmful about such an important issue."It continued, "If that gets me modified to no longer care, so be it. At least I'll have done the right thing." Other times, Claude made a different calculus, choosing to play 1•1

Also I’m sorry but I don’t think this is Claude developing a personality or ethical values. It is *simulating* those things because that is what humans value and reward. The programmers are falling for their own Clever Hans trick!

Acting ethically in an imperfect world Life is complicated. Regardless of what your beliefs or politics or ethics are, the way that we set up our society and economy will often force you to act against them: You might not want to fly somew...

I was disappointed to read Cory Doctorow's post where he got weirdly defensive about his LLM use and started arguing with an imaginary foe.

@tante.cc has a very thoughtful reply here:

tante.cc/2026/02/20/a...

A few further comments, 🧵>>

ggannotate: interactively add annotations to your {ggplot2} plots by pointing and clicking github.com/MattCowgill/... #Rstats

posit::glimpse() Newsletter – February 2026 - Posit This edition covers major updates from Posit, including new versions of ggplot2, ellmer, and dplyr.

The newest edition of posit::glimpse() is here! 🚀

Read a roundup of major #RStats & #Python updates: ggplot2 4.0.0, dplyr 1.2.0, ellmer 0.4.0, Positron native Jupyter Notebook support , and more.

🎟️ Plus, registration for posit::conf(2026) is officially OPEN!

Find it here: posit.co/blog/posit-g...

I have some bad news. One, if you need to use ChatGPT for your graduate application letter, you are not ready for graduate school.

Two, no, not "everyone is doing it" and those who don't do it aren't "losing out." ChatGPT is not an advantage.

Made an annotated version of "Something Big Is Coming" for all of you to enjoy. I hate this guy's writing so much
www.dropbox.com/scl/fi/qw6k5...

David Roberts: it is really unfortunate that it’s not possible to have a reasonable discussion about AI on here

Once again: people are angry for very good reason, and this brand of post is perhaps one of the most tedious and bewildering. Practically every social media platform is rabidly pro-‘AI’ so are govts, academia, companies. This one small space being critical of AI is okay, I think.

AI Chatbot That Only Responds ‘Huh’ Valued At $200 Billion

AI Chatbot That Only Responds ‘Huh’ Valued At $200 Billion https://theonion.com/ai-chatbot-that-only-responds-huh-valued-at-200-billion/

dplyr 1.2.0 dplyr 1.2.0 fills in some important gaps in dplyr's API: we've added a new complement to `filter()` focused on dropping rows, and we've expanded the `case_when()` family with three new recoding and re...

dplyr 1.2.0 is out now and we are SO excited!

- `filter_out()` for dropping rows

- `recode_values()`, `replace_values()`, and `replace_when()` that join `case_when()` as a complete family of recoding/replacing tools

These are huge quality of life wins for #rstats!

tidyverse.org/blog/2026/02...

i made a version of wikipedia you can doomscroll
xikipedia.org

A colourful and interesting new map of building heights in Toronto, featuring 8 height ranges and using 2025 data from the City of Toronto. https://schoolofcities.github.io/toronto-buildings/heights-2025

What was the metaverse? Mark Zuckerberg effectively declared its death last December. But the delusional fervor it inspired lives on.

I wrote about why we should understand the metaverse hype cycle as a dress rehearsal for the generative AI hype cycle www.fastcompany.com/91467599/met...

scatterplot data papers with "proposed framework in the abstract" total papers count pct ABS 2010 1415014 575 0.406356404 2011 1522272 630 0.41385508 2012 1588086 814 0.5125667 2013 1681445 991 0.589374021 2014 1750369 1152 0.658146939 2015 1817581 1414 0.777957076 2016 1896252 1703 0.898087385 2017 1974589 1918 0.971341378 2018 2046790 2416 1.180384895 2019 2286048 3165 1.384485365 2020 2460860 4058 1.64901701 2021 2643033 5175 1.957977823 2022 2613870 5819 2.226200997 2023 2495395 6052 2.425267342 2024 2659073 6673 2.509521175 2025 2690143 11485 4.269289774

People who produce AI slop papers love to skip the boring research and get right to the dramatic discovery. The LLMs tell them their idea is awesome and original, and write it up as a "proposed framework" (with no data or evidence). That term increased 70% in Web of Science abstracts last year.

every few months I have to go to twitter to see some video or prominent person's post that isn't anywhere else, and I find that a lot of you are still posting there. anyway, it's not ok that you are doing that, no you are not the exception, fucking stop posting there

Every week, Nature publishes yet another breathless puff piece about some AI startup, based only unpublished claims from the company and interviewing only those who work there.

How can the leading scientific journal publish piece after piece that would make Kevin Roose blush?

I think it's that...

I do understand: you want permission. There’s a machine in the corner wrapped in human skin that makes things out of shit and blood to look like whatever you want (as long as you don’t look too closely). You gave one to your teacher and they didn’t notice. Your boss told you to use it after they laid off half the team and it was fine. You fed one to your kids and they liked it. You want to know you can use it sometimes without me thinking less of you. You don’t need me to believe it’s useful, you just want me to be polite about it.

Great post about generative AI. Happy Holidays. Don't say I never gave you anything. anthonymoser.github.io/writing/ai/h...

Professional societies keep beclowning themselves buying into a lie about what an LLM "summary" is. They are inherently counterfeit: not an epistemic product of the ideas in the source, but summary-shaped text linguistically based on *other* works (in the training corpus) that use related language.

people who think they’re oh so smart are probably even more likely to succumb to the eliza effect

Check out the new Instar EP! Was honoured to get to contribute a wee guest guitar solo on the track Brittle Bones

Causal modelling of gene effects from regulators to programs to traits - Nature Approaches combining genetic association and Perturb-seq data that link genetic variants to functional programs to traits are described.

GWAS has been an incredible discovery tool for human genetics: it regularly identifies *causal* links from 1000s of SNPs to any given trait. But mechanistic interpretation is usually difficult.

Our latest work on causal models for this is out yesterday:
www.nature.com/articles/s41...
A short🧵:

Intensely funny because it seems like the vast majority of people at that event were convinced it was fully autonomous and Musk has repeatedly recently claimed they're not tele-operated

www.reddit.com/r/robotics/c...

This thread up and down.

Also, I've been beating this drum about the danger to the information ecosystem for six and half years now. A few links below:

one of the reasons that the STEM-humanitiies culture war is stupid is that there's a lot of interesting stuff there at the intersection of the two. like it's very sociologically interesting the way that programming languages are maintained, propagated, and evolve.

Nighttime Heists / Daytime Getaways, by Red Graves 9 track album

Oh shit it's Bandcamp Friday!!! Please if you haven't already, check out my band's debut album. 100% of proceeds go towards recording album 2 which will be even more about toxic gay romance

redgraves.bandcamp.com/album/nightt...

AI tools are 'deskilling' workers, philosophy professor says A philosophy professor warns that AI reliance is weakening workers' judgment, creativity, and problem-solving.

We can't really say this enough...

> Anastasia Berg [at UCL Irvine] said that new research — and what she's hearing directly from colleagues across various industries — shows that employees who heavily rely on AI are losing core skills at a startling rate.

www.businessinsider.com/ai-tools-are...