Notices by tante (tante@tldr.nettime.org)

Companies are realizing that pushing people to using "AI" is expensive (even at the subsidized pricing going on right now)

(Original title: Companies go full AI — then the bill comes due)

https://pivot-to-ai.com/2026/03/19/companies-go-full-ai-then-the-bill-comes-due/

I personally consider "I asked ChatGPT to generate a response to you" not witty but a form of an insult. Don't do that please. If that is how you want to talk to people at least don't tell me. It's offensive.

RE: https://hachyderm.io/@dalias/116286540925560722

That would just be one user account. The service connects with each user's account through a browser making it a bit annoying to detect. So you can block the developer but anyone else could use the same tool to scrape your posts and feed it into a slop machine.

OpenAI kills their Sora Video generator.

Which isn't that surprising TBH:
- it is massively expensive to run
- the results were not that usable
- it was a lawsuit generator

So in order to keep the numbers looking better till their IPO at the end of this year they are starting to cut shit. Just an indicator that things are not looking good cashflow-wise.

RE: https://tldr.nettime.org/@tante/116278629466465183

Spending 3 to 5 USD for every USD you make is not exactly what I'd consider good business.

"Microsoft 365 has more than 450 million paid commercial seats. After roughly two years on the market, Copilot has converted approximately 15 million of them into paying users. That's a 3.3% conversion rate, at $30 per user per month, generating roughly $5.4 billion in annual revenue. That's less than what Microsoft spent on infrastructure in a single quarter (3)."

None of the numbers make economic sense.

https://finance.yahoo.com/markets/stocks/articles/nadella-paid-650m-recruit-ai-170000361.html?guccounter=1

By ignoring the "master" part in the quote it is reduced to something one can cutesily brush aside: But it's not about appropriating someone's screwdriver. It's about whether you think you can appropriate his torture device.

Audre Lorde's "The master's tools will never dismantle the master's house." is not just a statement about a tool being tainted by its origin. It's about what kind of tool a "master" would create: Whips. Chains. Violence.

That's the meaning: You cannot just take tools whose purpose and politics is dominance and violence and "make them liberatory". This goes deeper than "just" embedded politics or lofty talks about ethics, it comes down to what kind of relations you believe do and should and must not structure the world.

RE: https://tldr.nettime.org/@tante/116239395175554368

So yeah. I am kinda looking for a new job that allows me to make our digital world meaningfully better. If you have any leads or thoughts, ping me.

RE: https://social.coop/@cwebber/116217477822586442

The whole "AI" industry is just based on the hope that they can deskill people fast enough that they'll have to rent back cognitive support systems. Forever.

It's like Uber. Just that they don't try to break existing transportation infrastructures but your brain.

Which was really fucked up to see: These folks actually want to protect their organizations from burning a lot of resources on bullshit instead of fixing actual problems that help the workers _and_ the organization. And they have to actively fight management who got their brains ruined on linkedin.

But: If you have any chance to speak to unions/workers from different domains and organizations do so.
It's fascinating how
a) different organizations are and operate
b) they all end up with the same handful of structural problems

But it was super fun to lead them through a "this is how you can force reasonable evaluation on 'AI' projects which kills most of them" framework and see how they felt empowered and able to actually do their job again.

Spend the day talking to workers council members about "AI". And it's kinda wild hearing their stories from the wild: Management is 100% in fantasy "AI" can do everything land and makes huge plans for how to use "AI" to cut workers when real projects that supposedly can do 50% of a specific task end up being able to do 8%. And they still go live. It's fucking bonkers. CEO's are really not okay.

"You go to sleep, the agent does the science and when you wake up, you have the results."

That's not how any of this works. You're not "doing research with help" you have something generated that looks like a research report. Without the research happening. How many different ways can "AI" bros find to express "I AM TOTALLY MISSING THE POINT"?

"AI" skills are basically just documentation and it says something about how documentation only gets resources and respect if it is framed as "technology". Which has gender bias all over it.

This case shows how Open Source will die. With anyone just being able to pipe existing code and tests through an LLM and claiming that to be "clean room" (which is hogwash) no licensing can protect your work from being accumulated and monetized by anyone. The commons are actively being shredded in front of our eyes.

https://github.com/chardet/chardet/pull/322

Yesterday I read an interview with the singer in a punk band who uses GenAI to help with writing his texts.

Punk. The whole point of punk was that if you know how to play 3 cords badly and have something to say just fucking go for it.

Good rule of thumb: Don't sign a petition that Steve Bannon also signed.

The question is not whether you can create software using LLMs - you can (most software is just boring CRUD shit).
But you do pay a hefty price: In lowering quality (security issues, less maintainable), in skill decay in the people "guiding" the stochastic parrots, etc.

It's not "can 'AI's create software" but "are we willing to accept worse software running more and more of our lives?"