AI - Artificial Intelligence or Artificial Ignorance?

[Tia]

https://techxplore.com/news/2026-04-ai-alarm.html#google_vignette

"Godfather of AI" warns there's a 10 to 20% chance AI could seize control

Speaking during an interview earlier this month that was aired on CBS Saturday morning, Hinton, who jointly won the Nobel Prize in physics last year, issued a...

www.techspot.com

Geoffrey Hinton, one of the three legendary computer scientists who have become known as the Godfathers of AI, is once again warning that the rapidly developing and lightly regulated AI industry poses a threat to humanity

 

[jp10558]

I find the overall "hate" for AI to be very odd. It's especially thick on sites like Reddit. You can't even post and say you've used AI without getting completely attacked. I suspect other revolutionary technologies have done similar things throughout history. I am sure there were many people that didn't like cars when they first came along. Cars killed off much of the blacksmith industry when the horse and buggy went away. We see similar things with EVs and self driving. I looked it up and similar has happened with the telephone, calculators and the printing press. Mostly things we simply take for granted today.

There was a podcast called "Pessimists Archive" that went over like 75 episodes of these sorts of things, one per episode. It was entertaining and enlightening. Bicycles were pushed back against too. Lots of things we just shake our heads at today. It always tempers my "new thing bad" if I ever get that instinct.

Perhaps AI has a bigger chance to kill a lot more jobs than those other technologies did, but I expect it will create different jobs as most of those other technologies did throughout history. AI certainly makes me more productive and helps me produce a better quality result in the end. I used to be more negative on AI and still don't really think of LLMs as true AI, but it has been helping me and I use them every day now.

I think there are multiple reasons people hate AI, with various levels of being a reasonable concern. Much of it is still in flux because things change over time, maybe on a weekly basis still.

1. I think there's some general hate of "slop" - this isn't AI specific, but AI sure made it easier for massive amounts of almost no effort "content" to be vomited out on the Internet in lists, "blog entries", advertising, youtube and other videos, music and more. At least for me, I find AI content like this to be worse versions of the old "Here's a 20 minute video to show you a paragraph of text content" or "Here's 10 pages of your supposed life story, relatives stories, and your journey to creating this bread recipe, with about 6-10 lines of ingredients and 1-2 paragraphs of instructions that you actually wanted.
   1. And worse, there's now a much lower signal to noise ratio because of all this junk.
2. There's hate because of the environmental and economic impact from the frontier AI companies. These tools are expensive, they drive up everyone's electricity rates, they use up scarce water, they sometimes pollute badly because grid energy can't keep up with the demands. They drive up costs of computers, game consoles, cellphones, and the like too. This drives up the costs of every other online service when a server that was $15,000 in 2024 became $73,000 in 2025.
3. There's the perceived de-skilling issues, which perversely also make using AI more dangerous. If you don't know the rough bounds of reasonableness for an answer, even at like the scale of should 20*5 be 2-4 digits long level of roughness, then you can get answers to that sort of question of 10,000,000,000 and figure - computer must be right. Whereas if you'd ever made change, done math, used a calculator, etc you'd immediately say - oh, the AI screwed that one up.
   1. Of course, this one is the most like the pessimists archive sort of concern. The current "danger" that's "new" is the ability to offload a lot more of your thinking to a computer, but there's always been people that offload what someone might claim is "too much" of their thinking to others. I'm not sure yet that this is a real concern.
   2. Part of the danger here is at scale - if you "trust" AI to do all your product research and store selection for you, at least currently the commercial ones from Google also point you towards ~30% more expensive options than Google's traditional search. Quelle surprise, right?
   3. AI is also at that "dangerous point" IMO of current "Self Driving Cars" - that is, it's good enough much of the time for you to trust it, but the vendors also say you should NOT USE IT without strict supervision. I.e. you need to be ready to take over for the self driving car at any time - but at that point you're basically driving anyway. Same with AI - if you have to double check everything, does it save you time? It makes it very appealing to skip those steps, but then you get caught by hallucinations or the like.
4. There's the privacy and security concerns - not necessarily new, but they surface lots of things that weren't secured, but also weren't noticed before. They make a lot more things actionable in a way that weren't before, for good and ill. I.e. automation in general has taken lots of things that while perhaps technically illegal in the past were not actually worth it to investigate and made it easy - speeding for instance. Also, there's the dystopian part where the AI hallucinates that you're someone else and you're arrested even though you have a drivers license showing your real name - the cops and government say always thing their tool is so good and you must have fake ID rather than thinking - hey, maybe this tool isn't that reliable.
   1. I think a lot of our laws will need some re-thinking to explicitly include what used to be the "not worth a cops time" level. Like if you manage to mis scan at the self checkout and miss a soda bottle - that probably should not be shoplifting even if it currently is. More mens rea, as well as some allowance for slop - i.e. set speed limits by 95% percentile of drivers rather than 65mph but everyone "know's" it's really 80mph.
5. Companies are using it to make customer service worse yet again. It's not really a step change but just a bit worse than before. We already often had companies that just had no way to get in touch with them, we already had companies that gave strict scripts to their customer service phone people, who often were overseas over a horrible VOIP link with accents that made it even harder to understand sometimes. I think these changes mostly hit that there is ever less ways to get through to someone who both can understand you're calling because the issue you're having isn't in fact solved by the website or you wouldn't have bothered calling AND can actually take any action the website couldn't because of scripts.

That said, I think a lot of the complaints miss that the issue isn't AI but lazy use of the tool, or not yet realizing that while OpenAI and friends charge insane prices, there are now "generic" models that are much cheaper, and presumably use less resources to achieve that, as well as ever better models you could run locally as time goes on for just the hardware cost and consumer laptop/PC electricity use. That is, AI is getting pretty close to being the same "cost" to everyone and environment as a laptop, and those haven't yet destroyed the world.

 

[DeniseM]

We just need this guy:

 

[CalGalTraveler]

Do they have punched the car (sold to their own service department) many months/years ago ? When they do this, the warranty starts at that time and when they sell it later, they should discount it as a used car even if it's brand new.
I bought one of my cars like that 10 years ago. The dealer punched the car in January, plated it in July for a sale director, he made 1200 km and I bought it in September with ~9 months lost on the warranty but I saved 20% of the car price like a 1 year used car.

Yes. punched the car. That is the term they used. They punched it 2 months prior to our purchase. 4 miles with the factory plastic on it is new enough for me. Especially given the discounted price. With tariffs, oil and increasing costs the same 2026 model is now $20,000 more.

 

[CalGalTraveler]

re: AI. I use Claude every day. I cannot imagine doing work without it anymore. It is very helpful when I am dealing with tense exchanges at work or complicated legal issues, or even reading a contract and telling me what items to be wary of and how to negotiate.

 

[Patri]

re: AI. I use Claude every day. I cannot imagine doing work without it anymore. It is very helpful when I am dealing with tense exchanges at work or complicated legal issues, or even reading a contract and telling me what items to be wary of and how to negotiate.

Shouldn’t your brain tell you that? This indicates future generations won’t even have our levels of critical thinking skills.

 

[Brett]

Shouldn’t your brain tell you that? This indicates future generations won’t even have our levels of critical thinking skills.

I remember in the early 1970's we started using electronic calculators replacing doing arithmetic manually.

Yeah, I had some brain rot but it apparently didn't hurt in the long run

 

[CalGalTraveler]

Shouldn’t your brain tell you that? This indicates future generations won’t even have our levels of critical thinking skills.

I use it to work smarter, not harder.

I am not a lawyer and I don't have time nor legal expertise to read the minutae in a contract. AI has passed the bar exam. Not perfect. There are times when you really need a lawyer. But for day to day much cheaper and faster than a lawyer to understand the gotchas in a contract and determine whether to consult a lawyer. Why spend hours pouring over a contract, trying to understand how the law is actually applied and the risks, when AI can tell you in 5 minutes.