Post

"Yellow Paint' for your brain but the paint is Lead-tin Yellow and also often wrong.

Joke's on you, I was never able to persist at hard things anyway

note also that Figure 2 is from a replication or the study they did in Figure 1 where they controlled for some confounds (participant ability, differences between the computer interface in the two conditions)

Thankful that I’ve never touched it

Cigarettes hook you chemically. AI hooks you cognitively Same leash, different drug The question isn't whether to pick it up. It's who's holding the other end

Absolutely. Completely wrong information presented in a confident manner. Apparently, surveys indicated that users prefer a confident answer, even if incorrect. That seems to explain the public support for people who lie brazenly.

Cigarettes make you way cooler than AI.

Yes, the risk of “deskilling” is why I don't use LLMs/chatbots.

That's exactly how I feel about it too.

Its not unlike the ‘self driving car’ stuff. Tech that monitors the driver and tries to prevent them doing something silly has undoubtedly saved many lives (if rather annoying at times) Tech that actually does the driving HAS killed people.

Interesting point. I guess its about how you choose to use it. I haven’t changed habits because I’m still PISSED OFF when it catches me out spelling separate as ‘seperate’ But I like it for catching typos becayse I’m worse at typing that spelling. (And yes - I left those two in deliberately 😉)

I would like to add that every spellcheck tool I've used has gotten noticeably worse the past couple of years, right around when everyone started throwing AI at everything. It very frequently recommends that I turn my sentences into nonsense. The grammar check is much worse than the spell check tbf

It doesn't really seem to do that with things I've been unsure about and just offloaded to spell check. Years later I'm still unsure about some words. When I write one I remember being unsure about it before and finding a solution but I can't recall which way was right.

This has been my big thing - I want to become _better_ at stuff, not give up.

I can relate to that. I have enough personal pride in my professional output to shun AI instinctively. But it's not JUST about pride either. As experts, we can plainly see that AI is neither proficient nor reliable enough to fully replace us. If human expertise decays, art is just... lost.

Maybe it's the former debate kid in me. Idk. But I'm unwilling to risk dulling myself for a manufactured sense of certainty and the unearned confidence that comes with using generative AI.

I think there are many ways an LLM might make my life easier, and also that I would absolutely overapply it to other scenarios, to my own ruin.

I feel the same, and I also feel I have problems with dependency. So I'm very careful around my habits.

That’s a great way to put it ‘like cigarettes’, I’ve felt this for a while. I’m not convinced I’m less susceptible to getting hooked than the average person, but I don not want to become LESS ‘able’ in any possible way.

The researchers offered the phrase "productive struggle", which is exactly the thing. Sometimes you need to be stuck to get unstuck.

Oh sure I’m painfully aware of people like that. I’m just saying that I couldn’t ever picture myself doing it.

Very well put. I’m starting to see it and am being fairly militant. I’m not a complete luddite but if someone is merely generating work for me using an LLM - by definition they are redundant and I’m making that clear. Pull your mental weight or fuck off.

And in the places where I am not yet good at things, I don't want to deprive myself of the opportunity to get better at those things

Occasionally thought of my mum "and if your friends all jumped off a bridge, would you do it too?"

How does one make it that long without getting frustrated and throwing it out the proverbial window? Every time I have interacted with AI it has been an active hindrance to what I am trying to do/find. Worse than literally nothing. Especially with search engines and customer support.

I immediately took the stance of "playing with the machine trains the machine" and have avoided it for ethical reasons, glad this means I've managed to avoid the addiction element of it too.

begging people to have a look at the data and draw conclusions for themselves

Seeing this clickbait title is like watching the old disinformation pass the torch to the new disinformation.

I think it's okay to check what a word means or what a concept entails when you can't ask the author (better than to remain unaware) otherwise, totally agree - lately, I'm trying to use the calculator less, with a similar goal in mind

that's a very good policy. we underestimate the effect google has on our brains

youtu.be/OxOl2GEaEGg?...

I didn't quit smoking to pick up something else

Yeah, about that… finance.yahoo.com/technology/a...

100% agree this is my experience

and I've not got the time out the spoons to write up whole paragraphs every time when converts will either ignore them or get equally annoyed there's another mark who's not yet part of their cult

Many good ways to hate AI Why are the bad ones so popular

It's very oversimplified, but that is literally what AI does. It takes random noise and runs it through a filter that brings it closer and closer to what it expects a response to look like.

yeh

Honestly, smoking is probably better for you than using AI.

This is unfair to cigarettes

As far as Zuckerberg, Thiel, Musk and co are concerned everything is going to plan. A generation of people incapable of independent thought is their wet dream.

You could say the same thing about a search engine. Spending days at a library gathering information from different collections gives you real appreciation of the pre-Ai Google