GNU social JP
  • FAQ
  • Login
GNU social JPは日本のGNU socialサーバーです。
Usage/ToS/admin/test/Pleroma FE
  • Public

    • Public
    • Network
    • Groups
    • Featured
    • Popular
    • People

Conversation

Notices

  1. Embed this notice
    Fesshole 🧻 (fesshole@mastodon.social)'s status on Sunday, 01-Mar-2026 21:20:09 JST Fesshole 🧻 Fesshole 🧻

    Forced to use AI at work, hate it. Usage is monitored so I can't avoid it. I've set up mine to play a Sim City clone. It burns through tokens, making crap cities that quickly go bankrupt. Just been told that I'm getting a bonus for being a top user.

    In conversation about 4 months ago from mastodon.social permalink
    • Embed this notice
      Rich Felker (dalias@hachyderm.io)'s status on Sunday, 01-Mar-2026 21:32:48 JST Rich Felker Rich Felker
      in reply to
      • Alex M

      @alexmu @fesshole Ignoring that the story is probably made up because there's an audit trail that'd get them fired, if they're required to use the thing in their job anyway, there is nothing unethical about meeting the requirement in a way that protects themselves from the cognitohazard and deskillling and sabotages the boss's intent. Same harm to planet either way and that's on the employer.

      In conversation about 4 months ago permalink
    • Embed this notice
      Alex M (alexmu@social.vivaldi.net)'s status on Sunday, 01-Mar-2026 21:32:49 JST Alex M Alex M
      in reply to

      @fesshole that is genuinely horrific. Those tokens you burn are actual resources in the real world, like water and fossil fuels.

      In conversation about 4 months ago permalink
    • Embed this notice
      Rich Felker (dalias@hachyderm.io)'s status on Sunday, 01-Mar-2026 21:53:39 JST Rich Felker Rich Felker
      in reply to
      • funnymonkey
      • Alex M

      @funnymonkey @alexmu @fesshole Thanks. This is why I replied. Even if the story is fake, the principle it prompts is worth thinking about.

      In conversation about 4 months ago permalink
    • Embed this notice
      funnymonkey (funnymonkey@freeradical.zone)'s status on Sunday, 01-Mar-2026 21:53:40 JST funnymonkey funnymonkey
      in reply to
      • Rich Felker
      • Alex M

      @dalias

      This.

      It's SOOOOO easy to blame the people at the receiving end of broken systems of power and control.

      When people refuse to acknowledge that there are many situations with no good options, and a small number of less bad options, we get destructive absolutism that hurts people.

      See also: voting in the US

      @alexmu @fesshole

      In conversation about 4 months ago permalink

      Attachments

      1. Domain not in remote thumbnail source whitelist: www.this.it
        Progetti architettura e servizi tecnici per immobili
        Consulenza tecnica di architettura ed ingegneria per progettazione, ristrutturazione di immobili, pratiche edilizie, perizie. Investimenti, valorizzazione e trasformazione di immobili
    • Embed this notice
      Rich Felker (dalias@hachyderm.io)'s status on Sunday, 01-Mar-2026 22:27:53 JST Rich Felker Rich Felker
      in reply to
      • Alex M
      • Kenny

      @weizenspreu @alexmu @fesshole The slop provider has it even without the employer doing anything. And the account belongs to the employer.

      In conversation about 4 months ago permalink
    • Embed this notice
      Kenny (weizenspreu@chaos.social)'s status on Sunday, 01-Mar-2026 22:27:54 JST Kenny Kenny
      in reply to
      • Rich Felker
      • Alex M

      @dalias @alexmu @fesshole „because there‘s an audit trail that‘d get them fired“

      What information do you base the assumption on that the employer implemented an actual audit trail (which might be illegal in Europe if it captures the conversion content due to data protection laws)?

      Many LLM providers provide mere usage statistics because of the legal issues.

      In conversation about 4 months ago permalink
    • Embed this notice
      Drey (atlovato@mastodon.social)'s status on Sunday, 01-Mar-2026 22:30:04 JST Drey Drey
      in reply to

      @fesshole - Right on, Right on.

      In conversation about 4 months ago permalink
    • Embed this notice
      Rich Felker (dalias@hachyderm.io)'s status on Sunday, 01-Mar-2026 22:46:03 JST Rich Felker Rich Felker
      in reply to
      • Alex M
      • Kenny

      @weizenspreu @alexmu @fesshole A casual remark about why a story is likely fake does not call for a detailed explanation of intricacies, especially when this kind of software is dropping logs both intentionally and unintentionally all over the place and there is absolutely significant risk of getting caught.

      In conversation about 4 months ago permalink
    • Embed this notice
      Kenny (weizenspreu@chaos.social)'s status on Sunday, 01-Mar-2026 22:46:05 JST Kenny Kenny
      in reply to
      • Rich Felker
      • Alex M

      @dalias @alexmu @fesshole So it‘s all based on assumptions communicated as facts and conclusions are drawn based on these assumptions. Understood.

      In conversation about 4 months ago permalink
    • Embed this notice
      🥑 Yours Truly! Unruly 🇨🇦🇪🇺🇺🇦🌻 (unruly@mastodon.social)'s status on Sunday, 01-Mar-2026 22:48:41 JST 🥑 Yours Truly! Unruly 🇨🇦🇪🇺🇺🇦🌻 🥑 Yours Truly! Unruly 🇨🇦🇪🇺🇺🇦🌻
      in reply to

      @fesshole

      Raises the question. Is there any good way to use AI? Is there a way to invert it and turn it on itself, to use AI to destroy AI, to sabotage any ill it is designed to inflict, to undo or replace its propaganda effect, to use it to release the Epstein files?

      :mastodon:

      :mastodon:

      In conversation about 4 months ago permalink
    • Embed this notice
      Rich Felker (dalias@hachyderm.io)'s status on Sunday, 01-Mar-2026 22:53:45 JST Rich Felker Rich Felker
      in reply to
      • Alex M

      @alexmu @fesshole I didn't say they want to expose you to a cognitohazard or deskill you. I said it's reasonable and ethical to protect yourself from these things by diverting their irresponsible spending to something that doesn't harm you and keeps you employed.

      In conversation about 4 months ago permalink
    • Embed this notice
      Alex M (alexmu@social.vivaldi.net)'s status on Sunday, 01-Mar-2026 22:53:46 JST Alex M Alex M
      in reply to
      • Rich Felker

      @dalias @fesshole that's a myopic way to look at it. Most likely the employer doesn't care enough about you to want to deskill you by exposing you to "cognitoharazds". They're following hype and/or they're misguided. I'd be tempted to say they don't even have a strategy around the use of llms, but that's just speculation. So you're not really sabotaging the company by burning tokens. What you're doing is moral posturing that makes you feel better about yourself because you're the main character in your story and you're sticking it to the man. But that's a pointless exercise that only ends up sabotaging the cause you're supposedly fighting for (taking down capitalism, llm hype, late stage capitalism death cult). By silently complying, and using the tokens you're just bolstering their policy.

      In conversation about 4 months ago permalink
    • Embed this notice
      Rich Felker (dalias@hachyderm.io)'s status on Sunday, 01-Mar-2026 23:01:57 JST Rich Felker Rich Felker
      in reply to
      • Alex M

      @alexmu @fesshole Those are not more productive ways, they're more brain rotting ways. And even if they were productive, you have no moral duty to maximize productivity for your employer at the expense of your cognitive facilities.

      Yes, enployers are threatening to lay off employees based on low "AI" usage metrics. What rock have you been living under?

      In conversation about 4 months ago permalink
    • Embed this notice
      Alex M (alexmu@social.vivaldi.net)'s status on Sunday, 01-Mar-2026 23:01:58 JST Alex M Alex M
      in reply to
      • Rich Felker

      @dalias @fesshole is it really mandatory? Have you taken it up with the employer? How do your workmates feel about it? Have you talked to them? If the employer is threatening to fire you because you don't use llms then there are labour laws that can protect you (at least in the civilized countries).

      Even if you are forced to use llms, you can do it in more productive ways. Use it to save work if you can, and use the spare time to join/form a union, raise awareness with your peers on how this is damaging to their careers, do something that actually has a chance of changing things.

      In conversation about 4 months ago permalink
    • Embed this notice
      Em & future cats 🇺🇦🐈🏳️‍🌈 (em_and_future_cats@mastodon.social)'s status on Sunday, 01-Mar-2026 23:41:03 JST Em & future cats 🇺🇦🐈🏳️‍🌈 Em & future cats 🇺🇦🐈🏳️‍🌈
      in reply to

      @fesshole So…. “Usage” is monitored, but the loophole is “what you use it on is not” 😹
      At least you found a hack to “not really use it” 😹 kudos to you 😹

      In conversation about 4 months ago permalink
    • Embed this notice
      RBWells (rbwells@mastodon.social)'s status on Sunday, 01-Mar-2026 23:57:51 JST RBWells RBWells
      in reply to

      @fesshole ours did too. I just messed with it a little with pictures, then figured out it could do a better job of putting a PDF into Excel, than the native apps in Adobe or Excel, so once a month use it for that. My boss uses it to turn messages into AI speak, and uses Chat GPT at home, oof. I can't even say it is increasing productivity, all it's doing is making up for even more useless programs and a vendor with insufficient reporting. It's not perfect even at that one task.

      In conversation about 4 months ago permalink
    • Embed this notice
      🇺🇦 haxadecimal 🚫👑 (brouhaha@mastodon.social)'s status on Monday, 02-Mar-2026 05:33:06 JST 🇺🇦 haxadecimal 🚫👑 🇺🇦 haxadecimal 🚫👑
      in reply to

      @fesshole
      If I was forced at $JOB to use so-called "AI" for source code generation, I'd give it real work-related prompts, in case anyone is tracking those, then totally ignore the generated slop.
      That would obviously be just a waste of time, but actually using the output would be a BIGGER waste of time.

      In conversation about 4 months ago permalink
    • Embed this notice
      Niels Moseley (trcwm@mastodon.social)'s status on Monday, 02-Mar-2026 05:38:55 JST Niels Moseley Niels Moseley
      in reply to

      @fesshole I approve!!

      In conversation about 4 months ago permalink
    • Embed this notice
      houseofmates (houseofmates@mastodon.social)'s status on Monday, 02-Mar-2026 05:58:22 JST houseofmates houseofmates
      in reply to

      LMFAOO love the creativity with that

      In conversation about 4 months ago permalink
    • Embed this notice
      Jess👾 (jesstheunstill@infosec.exchange)'s status on Monday, 02-Mar-2026 06:11:34 JST Jess👾 Jess👾
      in reply to
      • Rich Felker
      • funnymonkey
      • Alex M

      I'm now kinda curious whether different queries that cost a credit cost different amounts of compute. The least malicious compliance would be to just ask lots of simple queries that don't take much compute but still cost lots of credits.
      @dalias @funnymonkey @alexmu @fesshole

      In conversation about 4 months ago permalink
    • Embed this notice
      Justin Macleod (justinmac84@mastodon.social)'s status on Monday, 02-Mar-2026 06:23:34 JST Justin Macleod Justin Macleod
      in reply to

      @fesshole This is a sick company

      In conversation about 4 months ago permalink
    • Embed this notice
      Daniel M. Reck (danielmreck@mas.to)'s status on Monday, 02-Mar-2026 08:07:30 JST Daniel M. Reck Daniel M. Reck
      in reply to
      • Rich Felker
      • Alex M

      @dalias If there is an audit trail, it would not shock me if it is being watched with the same intelligence and diligence as the leadership who would impose a requirement to use #GenAI. (That is, none.)

      If the boss is dumb enough to make that rule, they are dumb enough to miss a subordinate sidestepping the rule.

      @alexmu @fesshole #FAFO #tech #AI #slop #MaliciousCompliance

      In conversation about 4 months ago permalink
    • Embed this notice
      Rich Felker (dalias@hachyderm.io)'s status on Monday, 02-Mar-2026 08:07:30 JST Rich Felker Rich Felker
      in reply to
      • Daniel M. Reck
      • Alex M

      @DanielMReck @alexmu @fesshole Audit trail is just inherent to using a "tool" that's somebody else's service. Of course they record everything. That's valuable data. The boss isn't necessarily going to read it or even try to figure out how they would, but when the employee gets caught playing games at work, they're going to start digging.

      In any case it's pointless to argue over this when the even more obvious reason it's likely-fake is that "power fantasy sticking it to the man expressed as an anonymous confession" is a whole genre and most of these things turn out to be fake.

      In conversation about 4 months ago permalink

      Attachments

      1. http://digging.In/

Feeds

  • Activity Streams
  • RSS 2.0
  • Atom
  • Help
  • About
  • FAQ
  • TOS
  • Privacy
  • Source
  • Version
  • Contact

GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.

Creative Commons Attribution 3.0 All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.