• Leon@pawb.social
    link
    fedilink
    English
    arrow-up
    33
    ·
    1 day ago

    If ads are what it takes to kill this company then I’m all for it. Fucking cram that app full to bursting.

      • Leon@pawb.social
        link
        fedilink
        English
        arrow-up
        12
        ·
        1 day ago

        I can’t wait for it to die off. My editor decided to spontaneously fucking change my SQL query for me the other day. It made up utter nonsense, completely unprompted. Fucking hate all this garbage.

        • Peppycito@sh.itjust.works
          link
          fedilink
          arrow-up
          3
          ·
          edit-2
          19 hours ago

          I’m not sure what SQL query is, but I tried to use quickbooks help chat thing and it was changing my words as I typed and churning out gibberish. I thought I was having a stroke. I ended up having to type in a notepad and copy it into the chat. Fucking ridiculous.

          • Leon@pawb.social
            link
            fedilink
            English
            arrow-up
            5
            ·
            13 hours ago

            A SQL (sequel) query is a way to ask a database for information. A database consists of tables (think sheets in excel) where each table has rows and columns. Rows are individual entries, and columns are properties on each entry. So you could have a “booking” entry with say, a “date” and a “customer.”

            Excel and databases can be presented (and in many ways function) in very similar ways, as a developer the amount of times I’ve written software to shift work away from an Excel document into something more approachable and generalised… honestly it’s like most of what I do. Last year I developed an entire scheduling and follow-up application that was previously one person organising a bunch of people via an excel document. It was a fun project, the person was really fun to work with and super engaged with things.

            Anyway, a query is almost plain English.

            SELECT Customer, Date FROM Booking
            WHERE Date BETWEEN  '2025-11-15' AND '2025-11-30'
            ORDER BY Date ASC
            

            This is just an example, what I did was a little bit more complicated, but not by a lot. The LLM just decided to randomly change the verbs and add in random columns that didn’t exist. Completely unprompted, almost like autocorrect gone haywire.

            Between the environment being a slow mess, and the LLM screwing me over, what would’ve normally taken ~5-10 minutes took over an hour. I don’t see how it’s supposed to improve productivity.

      • BennyInc@feddit.org
        link
        fedilink
        arrow-up
        7
        ·
        24 hours ago

        Next step will be the ads targeting the teacher instead of the student. Need to offer them good deals on their school supplies, if they have to buy that themselves. Also, can’t support them directly. AI wants to replace them anyway. Money machine needs to go Brrrrrrr…

  • django@discuss.tchncs.de
    link
    fedilink
    English
    arrow-up
    11
    ·
    1 day ago

    This time, OpenAI data engineer Daniel McAuley came to the firm’s defense, arguing that “this is not an ad (there’s no financial component).”

    However, McAuley admitted that the jarring experience leaves plenty to be desired.

    “It’s only a suggestion to install Peloton’s app,” he wrote. “But the lack of relevancy makes it a bad/confusing experience. We’re iterating on the suggestions and UX, trying to make sure they’re awesome.”

    Straight up gaslighting, I see.

  • NotMyOldRedditName@lemmy.world
    link
    fedilink
    arrow-up
    4
    ·
    edit-2
    1 day ago

    So first off, ads suck.

    Second, I could put up with ads if they are fully disclosed as ads with a sponsor tag or something, but if they start putting it into the conversation as if it wasn’t an ad, I’ll stop using it right away.

    • xyguy@startrek.website
      link
      fedilink
      English
      arrow-up
      9
      ·
      17 hours ago

      This always seemed like the ultimate goal of a chatgpt or a Gemini.

      Attract: make a service that magically gives answers, killing the old ways to find or research information.

      Extract: once everyone trusts AI implicitly, weave in advertising without anyone’s knowledge, allowing for extremely subtle and effective brainwashing.

      • 🍉 Albert 🍉@lemmy.world
        link
        fedilink
        arrow-up
        4
        ·
        12 hours ago

        doubt any ad blocker can do that. it isn’t like a pop ad or a banner, it’s more like the conversation recommends you sponsored products. like the normal conversations in the Truman show