all 18 comments

sorted by: hot top controversial new old
[–] 28 points 1 month ago* (4 children)

This is right on time. I read a discussion on how "Chinese models" are censored on topics like Tiannamen square. But the real issue is that we (average person) are only able to spot such things from a distance (either geographical or temporal.)

This means that "our" models may already have blindspots, reality twists, and the like. Only that we are not (on average) spotting it (yet)

Personalization of answers and bubble reinforcement will come. The algorithms have been increasing their capacity to influence us. Not only for their sheer power solving benchmarks but by the legitimacy that's being accumulated in the process.

However,

"ChatGPT says to stay calm and keep working hard" /s

  • source
  • hideshow 4 child comments
  • [–] 7 points 1 month ago* (3 children)

    I like to go ask them questions I know the will lie about. It helps to reenforce that they are untrustworthy.

    Anything about trump always come out as 1984 new speak.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 3 points 1 month ago (2 children)

    I felt the Impuls to answer to you: „just don’t converse politics with them.“

    Then I realized the issue goes deeper and beyond politics and is mostly very subtle to an oblivious user. Reality is narrated differently from country to country.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 1 point 1 month ago

    Well, while many of these chatbots have intentional knowledge injections regarding some political topics - they’re also simply suspectible to eating up disinformation from internet too. Most of major LLM chatbots were reported to cite shit like Pravda or RT. So yeah, you either get disinformation that is hardcoded into them or regurgitated from web.

  • source
  • parent
  • [–] 24 points 1 month ago* (last edited 1 month ago)

    Oh so true.

    Censoring gaslighting manipulative program forcing a corporate script into your brain.

  • source
  • [–] 9 points 1 month ago (3 children)

    These companies already made it obvious that information is something they want to control. They add their filters, that are enforced and decided by them. They are buying and destroying books to gather information and eliminate the chances for competitors to gather. They are hoovering up all the information they can and can manipulate it however they want.

    The game right now is to be the AI option with the most stolen information and curated how the investors want it to be.

  • source
  • hideshow 3 child comments
  • [–] 7 points 1 month ago (2 children)

    It's not just information. The conference I just attended was all about agent to agent interactions. They want a local model on everyone's desk, talking to a customer relationship manager agent that's going to live in their data centers or the cloud. It might run on Claude or something, but only until they can bring it in house to reduce their token costs.

    You won't even know what company you're dealing with. You'll just ask your agent for groceries, a taxi, whatever, and it shows up at your door.

    Then, once they control that relationship, they own you. Entrap and enshittify. Hopefully after you've stopped knowing how to think for yourself and can figure out how to break out and look for a competitor.

  • source
  • parent
  • hideshow 2 child comments
  • The narrative has always been someone else's, digital or analog.

  • source
  • [–] 2 points 1 month ago (3 children)

    Substitution check:

    It is time to understand without complaint that systems designed to enable computers to react to external inputs without need of human interaction are a list of events owned by someone else?

    Nope. Still just as incoherent. Fuck the LLM companies and the business monsters pushing them, but that second panel is less sensible as a sentence than the nonsense put out by early versions of chatGPT. At least that means it was probably written by a human, though, so that's nice.

  • source
  • hideshow 3 child comments
  • [–] 5 points 1 month ago (2 children)

    Hi, human who wrote that meme here. I often sucks at words. What I was trying to say is that LLM companies tend to heavily maniuplate the raw output of their models to fit their narrative, all whilst selling it as an independently thinking entity whose inner workings no human can understand.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 3 points 1 month ago

    Your prompts are not without system prompts.

    When they complain that Chinese models will do propaganda on you, what they really mean is “we didn’t get to put our propaganda into them”

  • source
  • parent
  • [+] 3 points 1 month ago (1 child)
  • [–] 2 points 1 month ago

    more like everyone’s

  • source
  • 16 comments in c/memes@lemmy.ml view that post
    [–] 8 points 1 month ago

    Ostensibly it’s the amalgamated narrative of the model’s corpus, but corporations, most notably SpaceXAI/Grok, will often tune the narrative to their taste.

  • source
  • [–] 12 points 1 month ago (5 children)

    That doesn't make any sense.

  • source
  • hideshow 5 child comments
  • [–] 1 point 1 month ago (2 children)

    Not sure to what extent this is true, but people say AI is made available for free because the profit comes from influencing people's views and thus cultivating profitable behaviours.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 7 points 1 month ago

    Aaaand it's running on someone elses computer.

  • source
  • [–] 3 points 1 month ago

    The industry knew we were dumb enough to pay thousands of dollars for our dumb terminals. The people of the world deserve their tech fudalism, we asked for it.

  • source
  • [–] 2 points 1 month ago

    if you refuse to think for yourself, someone else will do it for you.

  • source
  • [–] 2 points 1 month ago (4 children)

    It's actually lots of people's copyright-protected material all chopped up into a slop-smoothie.

  • source
  • hideshow 4 child comments
  • [–] 1 point 1 month ago (3 children)

    people’s copyright-protected material all chopped up into a slop-smoothie

    i like how this is prosed so i'm stealing it :p

  • source
  • parent
  • hideshow 3 child comments
  • [–] 3 points 1 month ago (2 children)

    as long as you're not a thieving LLM company, you may have it, for it is freely given

  • source
  • parent
  • hideshow 2 child comments