Login
You're viewing the functional.cafe public feed.
  • Nov 17, 2026, 7:19 AM

    this is a funny joke, but it's actually worse. the companies are going to find out that the codebases developed using LLMs are no longer human maintainable. Anthropic owns them now and can charge whatever it wants to shake the 8-ball again for a fix.

    Image attached toot
    💬 13🔄 39⭐ 0

Replies

  • Nov 17, 2026, 7:20 AM

    and Anthropic is going to be insolvent in a few years, so then what happens?

    💬 3🔄 0⭐ 0
  • Nov 17, 2026, 7:23 AM

    i think this is going to be the basis on which the capital class argues for some kind of bailout. corporate America is nearing the point where it depends on AI to do essential tasks, plus the top of the funnel for new CS talent is basically closed. they will argue that the government needs to fund inference as a public service to keep the economy running.

    💬 3🔄 3⭐ 0
  • Nov 17, 2026, 10:15 AM

    @peter I still expect them to manage to buy a few senators and judges to change the copyright law once they manage to get absolute dependence, so the chatbot is considered "sentient enough" for the company to own the copyright of the code made with it.

    💬 0🔄 0⭐ 0
  • Nov 17, 2026, 11:23 AM

    @peter
    "AI" as a public service?
    Well, that's an option I already thought about. But wouldn't that be too "socialist"?

    💬 1🔄 0⭐ 0
  • 💬 0🔄 0⭐ 1
  • 💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 17, 2026, 10:18 AM

    @peter when Anthropic goes under, so will all the companies reliant on them.

    💬 2🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 17, 2026, 7:33 AM

    @peter As foretold in the prophecies. By absolutely not listening to any experts, they brought this on themselves.

    💬 1🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 17, 2026, 7:47 AM

    @peter I was around in 1999, and people who knew COBOL made a fortune fixing code that no one else knew how it worked anymore.

    💬 1🔄 0⭐ 0
  • Nov 17, 2026, 8:06 AM

    @nahuellofeudo ah this is an interesting analogy: one hopes that they were able to retire with their riches or retool as they were essentially helping to phase out that codebase (mostly) - wonder if that's what's going to happen in the next wave of hiring for devs ...

    💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 17, 2026, 9:42 AM

    @peter Wait until:
    - Anthropic generates codebase, just to be hacked by the "ghost" of OpenAI...
    - People start training AI models for replacing CEOs and other executive suites, management boards, and possibly financial institutions...
    - When companies are unable to operate due to global AI hacking against all things digital...

    #digitalfreedom

    💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • bdf2121cc3334b35b6ecda66e471bdf2121cc3334b35b6ecda66e471@mastodon.social
    Nov 17, 2026, 10:24 AM

    @peter I think it's less likely that Anthropic charges these companies and the companies actually pay, and more likely that the companies just go under, since those companies would have to get that money from their customers and I don't think they are as necessary to their customers as Claude is to them.

    That was a terribly formatted stream of consciousness but I can't think well enough to fix it

    💬 0🔄 0⭐ 0
  • GhostOnTheHalfShellGhostOnTheHalfShell@masto.ai
    Nov 17, 2026, 10:28 AM

    @peter

    Nature red in tooth and claw.

    Such companies are destroying themselves with a form of self-inflicted cancer.

    Our job is to pick through the corpses and try to find those bits of software that haven’t been infected. Companies run as worker co-ops should be immune to the AI nonsense although I don’t think it’s a guarantee

    💬 0🔄 0⭐ 0
  • Nov 17, 2026, 10:59 AM

    @peter This is yet another case of #FAFO, and is good news for tech consultants.

    My sister is about to get even more job security due to these cheapskate companies going through the FO stage of the FAFO that is firing all the talented development and security teams, and replacing them with an #AI #Clanker

    When an #LLM #AI model faced with a long duration task or a very complex task, it tends to overload it's context window and that's when it starts to hallucinate like a clanker on bath salts.

    💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 17, 2026, 12:22 PM

    @peter it's neither good or bad.

    what you describe is a business decision made with all the necessary risk information readily available.

    mismanagement of a company is perfectly fine when the consequences of it affect the company.

    nobody owes us their company kept in business. they can make stupid mistakes and go out of business, who gives a shit. a company led by somebody who can think can and will just take their market niche over.

    💬 1🔄 0⭐ 0
  • Nov 17, 2026, 12:24 PM

    @peter for profit companies are not public institutions.

    it's only a bad thing when it is a government funded and run organisation providing essential services. which need to be run by serious people.

    for profit companies are the wild west! always have been.

    💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0