TechTakes

new episode of odium symposium. we look at rousseau's program for using universal education to turn woman into drones

https://www.patreon.com/posts/project-1789-150782184

Apparently some of our AI Safety cult "friends" are planning a protest in London on 28th of February.

Is it going to be something worth critically supporting instead of the usual criti-hype? Possible, but not likely.

The AI Safety movement is finally changing by ~~Sillyconversations~~ Siliconversations.

Who?

I used to be a quantum scientist and now I'm a YouTuber. My parents are thrilled.

Oh, okay.

Also curious that they're not protesting Anthropic on the thumbnail. A cynic would say they're giving them free pass because they say the right shibboleths.

They're giving them free pass because they say the right shibboleths.

What the fuck is going on with archive.ph, this is some 2000s internet drama shit

https://arstechnica.com/tech-policy/2026/02/wikipedia-bans-archive-today-after-site-executed-ddos-and-altered-web-captures/

What in tarnation is happening at adafruit?

https://blog.adafruit.com/2026/02/14/heres-our-first-gemini-deep-think-llm-assisted-hardware-design/

Somebody vibe-coded an init system/service manager written in Emacs Lisp, seemingly as a form of criticism through performance art, and wrote this screed in the repo describing why they detest AI coding practices: https://github.com/emacs-os/el-init/blob/master/RETROSPECTIVE.md

But then they include this choice bit:

All in all, this software is planned to be released to MELPA because there is nothing else quite like it for Emacs as far as service supervision goes. It is actually useful -- for tinkerers, init hackers, or regular users who just want to supervise userland processes. Bugs reported are planned to be hopefully squashed, as time permits.

Why shit up the package distribution service if you know it's badly-coded software that you don't actually trust? 90% of the AI-coding cleanup work is going to be purging shit like this from services like npm and pip, so why shit on Emacs users too? Pretty much undermines what little good might come out of the whole thing, IMO.

need a word for the sort of tech 'innovation' that consists of inventing and monetizing new types of externalities which regulators aren't willing to address. like how bird scooters aren't a scam, but they profit off of littering sidewalk space so that ppl with disabilities can't get around

EDIT: a similar, perhaps the same concept is innovation which functions by capturing or monopolizing resources that aren't as yet understood to be resources. in the bird example, we don't think of sidewalk space as a capturable resource, and yet

Chatbots are a cognitive hazard, part infinity: AI Delusions Are Leading to Domestic Abuse, Harassment, and Stalking

Russ Wilcox is not impressed by the Mass AI bill:

https://russwilcoxdata.substack.com/p/i-read-every-line-of-massachusettss

Four: create a private right of action. Let deepfaked candidates sue. Give them access to injunctive relief and takedown authority. If someone fabricates your face and your voice to destroy your campaign, you should be able to walk into a courtroom.

Hell yeah we need this.

ai;dr https://futurism.com/artificial-intelligence/aidr-meaning

Last megathread I posted about some LWer writing about "demographic collapse" in Japan (indistinguishable from anyone with any ambition leaving for greener pastures). I've read the comments to see if there's any pushback, and found this absolute doozy

https://www.lesswrong.com/posts/FreZTE9Bc7reNnap7/life-at-the-frontlines-of-demographic-collapse?view=postCommentsNew&commentId=Gbfu5BGaZMwXFey7h

this article that tries to argue that we have managed to achieve AGI already is a real hoot

Somebody on bsky talking about various Ben Goertzel Epstein file emails.

Starting the week with this fairly extensive compendium of intractable issues with LLMs: GenAI has an Alignment Problem… But it’s not the machine’s hostility we need to worry about.