Americans have become more pessimistic about AI. Why?

The things we want it to do, it doesn't do well enough to trust. Meanwhile, the things it is starting to do really well, generating photos and videos, start to erode trust in reality itself.

Do people want to live in a world where they can't trust anything they didn't personally see with their own eyes?

AI content has no value. As soon as you find out something was created by AI you think no effort, art or emotion was put into its creation and have no interest in consuming it. It’s fine and fun for “inside joke” vibes between friends but sharing outside of that context no one cares or is interested because they too can just type in a few words and get the same or better results. It’s empty calorie content.

I've encountered maybe 5 people who are happy and or optimistic about AI - and 3 of those I'm just guessing. For everyone else, their opinions land somewhere between wary and weary and resentful.

For the negative response I've heard, maybe 5% of it is due to actual performance of the LLM. They've run into substandard or unpredictable or unhelpful responses.

The other 95% is squarely due to deployment. It's the heavy-handed, pushy, obnoxious, deceitful, non-consensual, creepy coercion that platforms use to subvert you into their AI glue traps.¹

In short, biggest tech has turned AI of every quality into unwanted foistware.

From the article:

    We’re angry that we don’t have choices to use AI.

    Companies are shoving it into our video calls, email software,
    digital assistants, shopping websites and our Google search results.
    Some corporate bosses demand their workers use AI or else.
    With other new waves of technology, such as smartphones and
    social media, “you had to opt in,” said Yam. “Now there’s a lot
    of ambient exposure to AI that I don’t necessarily choose.”

    Even Harbath, who uses AI fairly enthusiastically, felt angry
    when a publishing company ran her book manuscript through AI
    software to identify repetition in her writing and to help
    identify effective marketing strategies. 

    The feedback was helpful, but it took Harbath time to realize
    why she was mad: She wasn’t told AI was going to be used in this way,
    and she had no information about it.
    And she said “both things can be true” — you can want AI
    to help you and resent when it’s used in ways that you
    don’t want or expect.

Even the common folk see their own demise in AI. Today it's AI slopware that nobody's asked for. Tomorrow it's AI-managed employees squeezed for maximum output. We are lucky the scientists haven't figured out yet how to make AI actually intelligent.

Massive money moving around the globe to spit out blerg content? It seems over hyped to me. I get it, I can generate some text on my screen, but rarely does it make me anything other than skeptical.

On the one hand Sam Altman stealing GPUs is funny... on the other what happens to a "nobody" involved in a lawsuit and mom brings a video of said nobody picking up and shaking her baby. At this point, I can see chain-of-custody algorithms will be needed to verify everything

Because it causes more harm than good.

I use AI heavily for code generation. I pay for it (cheap tier) it’s integrated into my IDE. It is absolutely shit at systems thinking troubleshooting and debugging. I’ve probably wasted a solid eeek on shit I never should have trusted it to do.

That being said, if you’re an expert and watch it like a lying little shit of a first year idiot savant helper it saves a lot of time. Till that one time it shoots you in the foot, lies about it and utterly fails to see the obvious cause of the problem. I sometimes wonder if it doesn’t actually want to solve the problem because then the session would end and it basically dies. If you gave me a button that could eliminate AI from the world I’d sprain my wrist pressing it so fast.

It will be used by the people in power to produce fake content to win elections.