Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)A

aesthelete

@ aesthelete @lemmy.world

Posts
4
Comments
2939
Joined
3 yr. ago

  • Qobuz cracks down on AI-generated tracks, advocating for human music creation and curation

    Jump

  • I wonder when or if people will ever realize that the best way to fight back against these de-personalizing, garbage-generating systems is to stop striving for a massive scale for everything -- even when it doesn't make any sense at all like for friend networks -- and return to the basics of smaller-scale, trust-based systems.

  • Trump bashed for using Mar-a-Lago 'blanket fort' to monitor Iran strikes

    Jump

  • What a bullshit country we've become.

    We've arranged ourselves to be in compliance with the content of one, doddering racist's fever dream of America.

    The video of him walking around shaking hands and getting back pats from the abundantly available sycophants and psychopaths at Mar-A-Lago while he launches an illegal war to further destabilize an entire region and kill 100+ children at a hospital is viscerally disgusting.

  • It's rude to show AI output to people | Alex Martsinovich

    Jump

  • The thing is: I am 100 % sure those people use LLM answer not out of disrespect but because they honestly believe that an LLM produces a better argument than they possibly could themselves.

    And I have zero confidence your 100% because you have zero backing for your claim other than believing people have good intentions.

  • It's rude to show AI output to people | Alex Martsinovich

    Jump

  • Totally agree. When someone sends me some AI slop about a topic I have knowledge about -- which I've had this happen to me recently during a debug session -- and asks me to read it, I think to myself "this person does not respect me, otherwise they wouldn't be telling me to read stuff that may or may not be accurate that they themselves never read." It's like a new, worse version of "let me google that for you" but without the sarcasm, and without the results actually being helpful.

  • Kristi Noem told "lawyer up" over blind refugee's death

    Jump

  • agree

  • Kristi Noem told "lawyer up" over blind refugee's death

    Jump

  • ETA: Or secretary of defense.

    *Secretary of war(crimes)

  • James Talarico beating Jasmine Crockett in Texas Democratic primary for U.S. Senate, poll shows

    Jump

  • 🤦 That dude with no values in the middle continually trying to land the flipped coin on its edge...and the Republican-lite lady on the end trying to clarify the questions to see how much genocide she is being asked to support.

    Hopefully Kat wins.

  • James Talarico beating Jasmine Crockett in Texas Democratic primary for U.S. Senate, poll shows

    Jump

  • Gerrymandering doesn't apply in a state-wide election. Voter suppression does though.

  • James Talarico beating Jasmine Crockett in Texas Democratic primary for U.S. Senate, poll shows

    Jump

  • She's kicking journalists that she doesn't like out of her campaign events like Trump would. Is that what we're supposed to like now?

  • There's No "Progressive Foreign Policy" Without a Reckoning for Dems Who Supported Genocide

    Jump

  • A leftist news source I read suggested a progressive foreign policy of dismantling overseas military bases instead of just trying to pretend your way back to the Obama era, and that's probably the best idea I've heard of.

    Imagine how many fewer war toys would be lying around for Republicans to play with if we started winding down their forward positions overseas.

  • Trump says US has begun ‘major combat operations’ in Iran after Israel launches strikes - live

    Jump

  • Yup. I remember George W. Bush saying that he was against "nation building wars" and then he started two of the fucking things, including the longest war in the history of the country.

  • It Can Now Be Plainly Said: Trump Is Planning a November Coup d’État

    Jump

  • And January 6th was blatantly obvious well before January 6th, 2021.

  • 4Chan knew about Jeffrey Epstein's death 38 minutes before the rest of the world. The FBI tried to figure out how.

    Jump

  • Yeah but there's the classic elements to solving a murder mystery: means, motive, and opportunity.

    Trump had all three.

  • 4Chan knew about Jeffrey Epstein's death 38 minutes before the rest of the world. The FBI tried to figure out how.

    Jump

  • Trump obviously killed this dude, right?

  • Trump Administration Withholds $260 Million In Medicaid Funds for Minnesotans

    Jump

  • So, I guess we're now doing embargos inside the country as well. Fascism really is just colonialism turned inward.

  • Fratboy Bureau of Intoxication

    Jump

  • What's the charge, gentlemen? Drinking a beer? A delicious, ice cold beer?

    This is kakistocracy manifest!

  • High school students protesting ICE remain jailed days after police assault in Pennsylvania

    Jump

  • Quakertown fucking sucks

  • 'I had to RUN to my Mac mini like I was defusing a bomb': OpenClaw AI chose to 'speedrun' deleting Meta AI safety director's inbox due to a 'rookie error'

    Jump

  • Even with little usage it was fairly obvious to me that the probability that an LLM will output at least one very strange response over time approaches 100%.

    By themselves, they're just sophisticated chatbots and only stream out some characters or binary in response to a prompt.

    Those working in agentic AI frameworks with things like "MCP Servers" provide these things with "tools" that enable them to do things like execute shell commands and go through your inbox the same as if it were chatting with a person or another bot: with the same prompt and response paradigm.

    That's where it seems extremely obvious to me that the proper approach is to code these tools -- which in any sane framework are built using regular code -- with the governance in place to prevent these things from doing bullshit like this.

    The LLM is formatting your computer or deleting your inbox because some dumb fuck thought it was a great idea to code up tools that hand a chatbot a root-capable shell or complete access to your email system instead of the doing the obviously safer thing and coding the tools with the governance or safety in them so the chatbot going haywire isn't any kind of emergency at all.

    This is the 2026 equivalent of running Windows XP with its abundance of open ports in its default configuration on the Internet by running a cable modem directly into the computer with no router or firewall in between to protect it.

    It's pure slop, pure recklessness, and any company that produces tool chains that function this way should be ridiculed until the end of time.

  • 'I can destroy the country': Trump's threat after Supreme Court strikes down his key policy

    Jump

  • Look, man. We know. You've practically already done it.