AI is Destroying Critical Thinking of People

Venusaur said:

I can already imagine kids just repeating like parrots what AI said, no source, no critical thinking, no other side, just repeating what their biased AI taught...and this bias is determined by people with very questionable goals and beliefs

Click to expand...

I remember reading 100s of replies below Twitter/X posts saying "grok, is this true?" lol

Nameless Knight said:

I remember reading 100s of replies below Twitter/X posts saying "grok, is this true?" lol

Click to expand...

I would personally support legislation that required you to be 18 to talk to AI. Instead of having AI companies gather your ID to verify you, however, I think it makes more sense to buy pre-paid codes from a gas station, where we already check for IDs for things like alcohol. We should acknowledge that an AI chatbot offering real-time, human-sounding emotional support is incredibly dangerous to a developing mind.

dealnotifier said:

I would personally support legislation that required you to be 18 to talk to AI. Instead of having AI companies gather your ID to verify you, however, I think it makes more sense to buy pre-paid codes from a gas station, where we already check for IDs for things like alcohol. We should acknowledge that an AI chatbot offering real-time, human-sounding emotional support is incredibly dangerous to a developing mind.

Click to expand...

Yes, but I also think that instead of using this technology to produce slop, the AI companies should focus all their business on helping healthcare, scientific research, and other such organizations (provided that AI can actually help them do their work)

If that happens, we might get a disease free society, clean air, and what not

Nameless Knight said:

Yes, but I also think that instead of using this technology to produce slop, the AI companies should focus all their business on helping healthcare, scientific research, and other such organizations (provided that AI can actually help them do their work)

If that happens, we might get a disease free society, clean air, and what not

Click to expand...

One complicating factor in all of this is China. Their open source models are getting very powerful, and since they are freely available on the internet, anyone can run them with the right hardware.

I've personally ran Qwen 3.6 on my rtx 3090 for small coding tasks, and it's as good as Claude Opus was in mid-2024 IMO. And that's just a model that can fit on consumer hardware. It was released earlier this spring.

I also annotate AI coding data for a living. After 3 years of doing that, I've concluded that the best way to picture AI is as a sycophantic calculator. Even if it's right a lot of the time, it's so eager to please you that you should respect it as much as any other kiss-ass in your life.

AI is perfect for analyzing or researching on subjects we have strong background in. To facilitate error detection, a user is expected to have some details about their needs, in this balanced level of conversation, AI is good, because it hates to argue, and accepts mistakes, when a user corrects a false or wrong answer, then in return it'll pop out the accurate details. The way humans use AI strictly illustrates how humans used humans before AI. humans find a way not to balance things, rather they prefer to transfer the whole workload on a person willing to assist. AI has no issue for users who understands what they're searching for, smart peole who detects errors.

airtman said:

AI is perfect for analyzing or researching on subjects we have strong background about, to facilitate error detection a user is expected to have some details about their needs, AI is good in this balanced level of conversation, because it fully accepts being wrong when a user corrects a false or wrong answer, then in return it'll pop out the accurate details. The way humans use AI strictly illustrates how humans used humans before AI, humans find a way not to balance things, rather they prefer to transfer the whole workload on a person willing to assist. AI has no issue for users who understands what they're searching for and could easily detect an error themself.

Click to expand...

Have you noticed that "vibe coders" get incredibly depressed when they run out of tokens? It's weird. How many SaaS products make you depressed like that when you hit your limits?

I think it suggests that amateurs see AI as a slot machine shaped to their whims rather than a research and implementation tool.

dealnotifier said:

I think it suggests that amateurs see AI as a slot machine shaped to their whims rather than a research and implementation tool.

Click to expand...

No Technology ever made by man promoted dumbness, AI was built under the foundation of High IQ people, and requires such level of input to work better. Calculators never made math easier for students who know no formulas, it was built to make computation easy for students who understands the right formula to solve a problem. AI works with prompt, a formula that determines what AI would say, users with shallow or zero indebt knowledge about a subject will get something shallow from AI, unlike when a pro or smart users, or industry guys with the right Jargons for a specific niche query AI, the results for an average or less informed user is not satisfying.

Stavro said:

Just like people manipulated newspapers, magazines, radio, and TV to spread misinformation and push an agenda.

The difference with AI is that the manipulation can be automated and scaled far more easily, but that's an acceleration of a problem that already existed, not something unique to AI.

Click to expand...

That's true. Rich & powerful people have always manipulated others to retain their status quo. AI has decentralised the system for anyone to pull the strings. Any simple action with the help of AI can have a snowball effect.

Somewhat off topic, but I have high hopes for AI's use in narrative RPG games like Dwarf Fortress, or mods of old games getting NPCs hooked up to LLMs. Since AI can call tools now, you could have NPCs give out custom items or quests in the game too. It sounds like a fun way to re-experience classic games.

I remember a project on github that simulated a village full of LLM-powered villagers, I wish I could remember its name.

I don't think its making people less critical in how they think if someone lacks strong critical thinking then they are going to do so regardless of Ai. They way you are describing it does not reflect lack of critical thinking either. Its just a way to aggregate Data fast. Want to search for the best vehicles prices or something and Ai prompt will aggregate a list with a median price while also producing some links to find those items.

On the flipside Ai will make people that already have a level of intelligence even smarter. I have used in several areas as sort of like a second opinion against my own and it has help me to polish my approach say i am making a thesis on how Thermal Dynamics works well i already have a base understanding but Ai would help me to formulate a more polished approach in my Thesis. Because its bouncing ideas back to me in something i understand it also help me to learn and pick up information and sometimes sharpen my understanding.

We know Ai is here to stay and atm its the leading industry that is going to dominate economically in many areas of society right down to physical infrastrucre in the real world. Can't beat em then join em i guess.