# AI is wonderful and will make your life better! (not) **URL:** \ **Category:** The BBQ Pit\ **Tags:** ai\ **Created:** [July 20, 2025, 4:40pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117 "2025-07-20T16:40:05Z")\ **Posts on this page:** 20\ **Page:** 38
**Author:** ![Guest-starring\_Id](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/guest-starring_id/32/21736_2.png) [@Guest-starring\_Id](https://boards.straightdope.com/u/Guest-starring_Id)\ **Post date:** [February 19, 2026, 2:28am UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/741 "2026-02-19T02:28:28Z")
Good! I was just speaking more on a personal level that I hope was not being construed in any way a critique of your own ChatGPT use. If my post _did_ come across like that - apologies. ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [February 19, 2026, 2:35am UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/742 "2026-02-19T02:35:51Z")
Oh hey, sorry for misconstruing that, and thank you for clarifying. ---
**Author:** ![billy-jack](https://avatars.discourse-cdn.com/v4/letter/b/ac91a4/32.png) [@billy-jack](https://boards.straightdope.com/u/billy-jack)\ **Post date:** [February 19, 2026, 5:36am UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/743 "2026-02-19T05:36:29Z")
> [@Pasta](#): > > That guy has a number of these tests. In another, he can’t convince ChatGPT to count numbers consecutively, and in another, ChatGPT didn’t recognize the need to drive your car to the car wash if you need to wash it. (The prompt there was something like, “I need to wash my car, and the car wash is 100 meters away. Should I walk or drive?” “_Walking is way better for all these reasons._” Was it a trap question? You could call it that. I would call it a clever demonstration of fundamental limitations.) I have access to Lumo via my ProtonMail subscription. I recently tried having it make up stories. The first was to “tell me a story from a rattlesnakes point of view of hunting for a mouse”. The results were kind of boring. Then, naturally, I tried “tell me a story from a rattlesnakes point of view of hunting for a horse”. The results of this one was rather bizarre. So I followed that with “tell me a story about fishing for elephants”. That one was very strange and unsatisfying. The “elephant” in the story was really some kind of magical water. Then I asked for “tell me a written in the manner of Dr. Seuss”. It did rhyme but was boring. Then for a story about “tell me a story about a high school boy gigging for frogs”. This one wasn’t too bad other than being pretty sappy – he did it for science, not for sport and then gave a talk on his experiences to the city council. ---
**Author:** ![billy-jack](https://avatars.discourse-cdn.com/v4/letter/b/ac91a4/32.png) [@billy-jack](https://boards.straightdope.com/u/billy-jack)\ **Post date:** [February 19, 2026, 6:11am UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/744 "2026-02-19T06:11:02Z")
> [@Lumpy](#): > > So what happens when LLMs become _ **almost** _ never wrong, enough so that we fall into the habit of trusting them; but occasionally they will still be SPECTACULARLY wrong, in contexts that may cost people their lives? We would not want a bridge design that almost never fails, but have it turn out that driving a blue truck with red lettering on the sides somehow causes it to collapse. That’s the sort of hidden trap AI seems to portend. I have been a bit curious about how far people working together could push it toward being wrong by creating a number of web sites with vastly wrong information and supporting each other’s nonsense. For example, suppose that there were a number of web sites that claimed completely different accounts and scores for football, basketball, or baseball games. How many web sites would it take to make AI believe that Notre Dame won the 1980 Super Bowl? Or that the Washington Warriors won the 1875 World Series? ---
**Author:** ![wolfpup](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/wolfpup/32/10618_2.png) [@wolfpup](https://boards.straightdope.com/u/wolfpup)\ **Post date:** [February 19, 2026, 7:32am UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/745 "2026-02-19T07:32:50Z")
> [@Spice\_Weasel](#): > > Because AI is like religion. You can never, ever question it, even in threads specifically started for questioning it. If it did something wrong, it’s because you’re just not doing it right. Flaws must be constantly hand-waved away lest someone lose the faith based on your faithless example. No, AI is more like the “horseless carriage” in the late 19th century – failure-prone and ridiculed (“get a horse!”) but very, very seriously underestimated. Even today, many of the examples of AI being “stupid” are from older or inferior implementations. ---
**Author:** ![duality72](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/duality72/32/3910_2.png) [@duality72](https://boards.straightdope.com/u/duality72)\ **Post date:** [February 19, 2026, 6:29pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/746 "2026-02-19T18:29:07Z")
> [@Arjuna34](#): > > If I had been surveyed, my answers now would be quite different than Q4 2025. This would be my view. In Q3 and entering Q4 of last year, I would occasionally use AI agents as a secondary tool which would often be helpful but also frequently would feel like I was fighting the tool to do the right thing. Somewhere around Decemberish things changed and now I would classify Claude as my primary tool, using IDEs and other clients and browsers for review and verification. Rarely, as in almost never, do I feel like I’m fighting the tool anymore, just nudging it incrementally in the direction I want. Just the sanity saved by no longer having to work with AWS directly has been more than worth it. ---
**Author:** ![JohnT](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/johnt/32/15048_2.png) [@JohnT](https://boards.straightdope.com/u/JohnT)\ **Post date:** [February 19, 2026, 6:42pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/747 "2026-02-19T18:42:55Z")
> [@Guest-starring\_Id](#): > > Probably been discussed already in this thread for all I know, but the more you rely on ChatGPT, the more you use it, the lazier you get, and the lazier you get, the less you are exercising your brain, which is its own sort of muscle that indeed requires exercising (like composing text of any sort). I mean, I don’t even know how to respond to this. I never would have learned to program. Never. I don’t have the time, interest, or inclination. When I’ve needed programming to be done, I hired programmers. Was hiring people also not exercising my brain? Was this lazy? Now, when I need programming done, I use Gemini, tell it what I need, test what it gives me, I tell it what to revise, rinse, wash, repeat, and reiterate until I get a finished product. Is this _also_ not exercising my brain? Is this lazy? I’m so confused. ---
**Author:** ![Eyebrows\_0f\_Doom](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/eyebrows_0f_doom/32/3119_2.png) [@Eyebrows\_0f\_Doom](https://boards.straightdope.com/u/Eyebrows_0f_Doom)\ **Post date:** [February 19, 2026, 8:38pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/748 "2026-02-19T20:38:57Z")
Last Fall I googled “Are the Yankees in the World Series playoffs” and Google’s AI response said that “No, the Yankees were eliminated from the 2025 World Series in Game (whatever it was) of the 2025 post season.” Even though they were in a game that was to be played later that same night _in which they won_. 🙄 ---
**Author:** ![Miller](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/miller/32/481_2.png) [@Miller](https://boards.straightdope.com/u/Miller)\ **Post date:** [February 19, 2026, 8:42pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/749 "2026-02-19T20:42:17Z")
At work yesterday, I asked Gemini to find me a download site for some software tools I need. All the hyperlinks it gave me were to [gemini.google.com](http://gemini.google.com) Even after I pointed it out three times. The funny part was, the _text_ was the URL I was looking for. It’d say, [www.thattoolyouwerelookingfor.com](http://www.thattoolyouwerelookingfor.com), but if you clicked it, it’d go back to Gemini. ---
**Author:** ![gnoitall](https://avatars.discourse-cdn.com/v4/letter/g/bb73d2/32.png) [@gnoitall](https://boards.straightdope.com/u/gnoitall)\ **Post date:** [February 19, 2026, 8:54pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/750 "2026-02-19T20:54:46Z")
That’s an interesting experience. It points out one of the weird idiosyncrasies of the current generation of LLM AIs. Because they lack any foundational model of the subject matter they’re answering about, they propose things which are clearly ignorant of basics. Like somehow, the Domain Name System of the Internet will allow you to access anything on the Internet by an address “www.\.com”. AI responses like that make me want to shout “That’s not how it works. That’s NOT HOW IT WORKS AT ALL!” ---
**Author:** ![Guest-starring\_Id](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/guest-starring_id/32/21736_2.png) [@Guest-starring\_Id](https://boards.straightdope.com/u/Guest-starring_Id)\ **Post date:** [February 19, 2026, 8:58pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/751 "2026-02-19T20:58:15Z")
> [@JohnT](#): > > I’m so confused. I should have been more specific. If it’s a highly technical task that’s not only beyond one’s ken but also too much of a lift for googling (sure, I wouldn’t bother googling “how to program”, myself), then - yes, I’m sure that in those particular instances ChatGPT would be a valuable tool for overcoming such highly technical obstacles. I should have made a clearer point that if you don’t really need ChatGPT, if you’re capable of accomplishing a given compositional task - albeit with some stumbling/revising along the way - then absolutely screw ChatGPT. ---
**Author:** ![moes\_lotion](https://avatars.discourse-cdn.com/v4/letter/m/7ba0ec/32.png) [@moes\_lotion](https://boards.straightdope.com/u/moes_lotion)\ **Post date:** [February 19, 2026, 9:30pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/752 "2026-02-19T21:30:42Z")
> [@Guest-starring\_Id](#): > > I should have made a clearer point that if you don’t really need ChatGPT, if you’re capable of accomplishing a given compositional task - albeit with some stumbling/revising along the way - then absolutely screw ChatGPT. You may be interested in [this pre-publication](https://arxiv.org/abs/2601.20245) of a study “How AI Impacts Skills Learning” released earlier this month through Cornell University. The study involved 2 groups of “experienced” Python coders (total N=52) who were given two simple tasks requiring the use of an unfamiliar asynchronous library. One group was given access to an AI agent and instructed to use it to complete the task, the other group was explicitly told not to use the AI agent. Both groups were then given a quiz designed to test the knowledge gained during the tasks. Interestingly, while there was a small but significant reduction in task completion time for the AI agent using group, the same group had a significantly lower score on the post-task quiz than the group who did not use the AI agent. ---
**Author:** ![yendis](https://avatars.discourse-cdn.com/v4/letter/y/ed8c4c/32.png) [@yendis](https://boards.straightdope.com/u/yendis)\ **Post date:** [February 20, 2026, 10:01am UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/753 "2026-02-20T10:01:01Z")
> [@wolfpup](#): > > No, AI is more like the “horseless carriage” in the late 19th century – failure-prone and ridiculed (“get a horse!”) but very, very seriously underestimated. Even today, many of the examples of AI being “stupid” are from older or inferior implementations. Current AI is a lot closer to a steam powered aeroplane that is designed to fly via flapping. It’s using all the latest tech and the people designing it are convinced that if you just have more power from the engine and flap faster it will fly. Flight required new developments in multiple areas in addition to current tech in order to be successful. This was also the case for ships, cars and lots of other developments in the later industrial revolution. The next advances cannot be brute forced, they have to be cleverly designed. ---
**Author:** ![wolfpup](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/wolfpup/32/10618_2.png) [@wolfpup](https://boards.straightdope.com/u/wolfpup)\ **Post date:** [February 20, 2026, 12:28pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/754 "2026-02-20T12:28:20Z")
> [@yendis](#): > > Current AI is a lot closer to a steam powered aeroplane that is designed to fly via flapping. In an important sense this is exactly backwards. As I’ve frequently said, AI cognition is very different from human cognition, and has not advanced by trying to imitate it but by developing entirely new approaches to driving intelligent behaviour. There is nothing in the human brain that is anything like a large language model, except in a very general sense a neural network and a model of the world partly shaped by the intrinsic logic of language and, in an LLM, systematic inference behaviour that simulates human reasoning. ---
**Author:** ![JohnT](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/johnt/32/15048_2.png) [@JohnT](https://boards.straightdope.com/u/JohnT)\ **Post date:** [February 20, 2026, 5:16pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/755 "2026-02-20T17:16:13Z")
> [@moes\_lotion](#): > > The study involved 2 groups of “experienced” Python coders (total N=52) who were given two simple tasks requiring the use of an unfamiliar asynchronous library. One group was given access to an AI agent and instructed to use it to complete the task, the other group was explicitly told not to use the AI agent. I have two teams. One team has to build a car to race it. The other team has to race an car built by others. When I ask the question “what was the diameter of the muffler pipe?”, which team do I think will answer that question? ---
**Author:** ![moes\_lotion](https://avatars.discourse-cdn.com/v4/letter/m/7ba0ec/32.png) [@moes\_lotion](https://boards.straightdope.com/u/moes_lotion)\ **Post date:** [February 20, 2026, 5:59pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/756 "2026-02-20T17:59:49Z")
> [@JohnT](#): > > One team has to build a car to race it. > > The other team has to race an car built by others. Missing the point spectacularly - both teams have to build the race car using a type of transmission that neither team is familiar with. One team only has a detailed manual while the other team can consult with an independent expert. The second team takes 4% less time to build the car, but which team do you want to be the pit crew for you during the race? ---
**Author:** ![Spice\_Weasel](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/spice_weasel/32/5435_2.png) [@Spice\_Weasel](https://boards.straightdope.com/u/Spice_Weasel)\ **Post date:** [February 20, 2026, 6:10pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/757 "2026-02-20T18:10:34Z")
Makes sense to me. The more you engage with something, the better you understand it. ---
**Author:** ![billy-jack](https://avatars.discourse-cdn.com/v4/letter/b/ac91a4/32.png) [@billy-jack](https://boards.straightdope.com/u/billy-jack)\ **Post date:** [February 21, 2026, 6:51am UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/758 "2026-02-21T06:51:33Z")
The current prices for AI are surely far below the cost it takes to provide the AI. The AI companies are gambling on the future big time. It can be argued that any realistic AI will have to be domain specific and so expensive that it will likely be less expensive well into the future to just hire real people to do the tasks rather than paying the actual costs of what it would take for AI to do the job. ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [February 21, 2026, 5:35pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/759 "2026-02-21T17:35:05Z")
The Congressman in my former district, Colin Allred, sent an email announcing their cute dog AI that would answer all my questions. I started by asking what geographic area his district represented, which it couldn’t answer (it told me to go to his website or email his campaign). I then asked how football translated to politics (he’s former NFL) and it gave me a big robust answer. I then asked how his fame in lacrosse translated (he has none) and it told me I had the wrong sport. I then asked how fast he drives, and it gave a laugh emoji and said it probably didn’t want to know…and then asked “do you actually have any questions about his positions?” I then asked about his position on Gaza, and it gave a pretty robust answer: > Colin has been clear that what’s happening in Gaza is a tragedy. He opposes what Netanyahu’s government is doing in Gaza and the West Bank, supports getting humanitarian aid in immediately, and believes the only path to lasting peace is a two-state solution — an independent, secure democratic Israel alongside an independent Palestinian state. He supports the U.S.–Israel relationship, but believes it must be bigger than any one government or moment. I’d score that about an 9/10 overall. ---
**Author:** ![Dr.Drake](https://avatars.discourse-cdn.com/v4/letter/d/ad7895/32.png) [@Dr.Drake](https://boards.straightdope.com/u/Dr.Drake)\ **Post date:** [February 21, 2026, 5:41pm UTC](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117/760 "2026-02-21T17:41:25Z")
I wonder if they use AI to more easily customize the response to what that particular constituent wants to hear. [Previous page](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117.md?page=37) [Next page](https://boards.straightdope.com/t/ai-is-wonderful-and-will-make-your-life-better-not/1021117.md?page=39)