[jplusequalt]

>I don't think "LLMs making us dumber" is a real concern. Yes, people will lose some skills. Before calculators, adults were probably way better at doing arithmetic.

But it's a problem of scale.

Calculators are very specific tools. If you are trying to run a computation of some arithmetic/algebraic expression, then they are a great tool. But they're not going to get you far if you need help understanding how to file your taxes.

LLMs are multi-faceted tools. They can help with math, doing taxes, coding, doing research, writing essays, summarizing text, etc. Basically anything that can be condensed into an embedding that the LLM can work with is fair game.

If you're willing to accept that using a tool slowly erodes the skill that tool was made for, then you should also accept that you will see an erosion of MANY skill you currently have.

So the question is whether this is all worth it? Is an increase in productivity worth eroding a strong foundation of general purpose knowledge? Perhaps even the ability to learn in the first place?

I would argue no a million times over, but I'm starting to think that I'm an outlier.

[dakolli]

Yeah, I agree. However, people use llms for the same reason people drive 3 blocks to a store rather than walk. Laziness and convenience. They simply don't care if their leg muscles atrophy. However, I think people aren't taking into account how much more important your thinking "muscles" are and its way more consequential to let those atrophy.

Everyone is vulnerable to the allure of taking shortcuts in life, but I've learned over the years that there is no free lunch. This is going too be quite an expensive trade off for many.

[slibhb]

People will have to be more intentional about using their increased leisure time in a healthy way. There was no point in exercising if you were a peasant who worked the field all day. Today, if you sit down in an office all day, you need to exercise intentionally. People have figured this out!

Along the same lines, AI will necessitate a shift where people intentionally use their extra intellectual leisure time. Reading, writing, chess, learning a new language, etc.

Not everyone will do this. Some people will be the intellectual equivalent of obese. But people will figure it out eventually.

[jplusequalt]

>But people will figure it out eventually.

Will they? >50% of US adults are overweight or obese. Is this the example you want to hang your hat on?

[slibhb]

People are figuring it out in real time. The next generation is going to be way less fat than the current one, because everyone exercises. It took time for people to adjust to a world where physical exertion is optional and delicious food is cheap, but we are getting there. I see no reason to assume the same thing won't happen with AI.

[jplusequalt]

>People are figuring it out in real time.

Where are the stats backing this claim? Obesity levels have not dropped significantly in recent times. Also, any significant change will require government oversight, and we are increasingly heading towards a direction where private interests overrule whats best for the public at large.

>I see no reason to assume the same thing won't happen with AI.

You have the ability to choose what and much you eat. Will you have the ability to forsake AI if your employer forces it upon you, or if to stay competitive in school you have to rely on it? In the same way it's hard to live in society without a smart phone, it's already becoming hard to operate in society without relying on AI. Now extrapolate this out by a decade.

I suggest you watch the AI Dilemma.

[slibhb]

The written word isn't a very specific tool. Before writing, people had to memorize things. In some sense, writing has made us dumber as memorization has been deemphasized. But was it worth the trade? Yes.

If you want a more recent example, google search is an extremely broad tool that has operated similarly.

I think AI will be another rung in the ladder of abstraction. Something will be lost, but it's worth the trade.

[jplusequalt]

I don't agree that writing, or Google search are on the same level here. A problem about having this argument on HN is that I think most people are already firmly entrenched in the pro-AI position, and will not consider any possible downsides.

[slibhb]

There are lots of anti-AI commenters on HN. Also, I didn't say there are no downsides. There are downsides to writing! And some people were against writing, like Socrates.

You should ask yourself why you're okay with innovations that happened in the past but not okay with innovations happening now. It could just be reflexive conservativism.

Of course there's no guarantee that AI will be more positive than negative, but I see no compelling reason to believe that. Most of the anti-AI sentiment is just people not liking new things.

[jplusequalt]

>You should ask yourself why you're okay with innovations that happened in the past but not okay with innovations happening now.

Because these innovations are not congruent with our most important biological advantage! We are here precisely because we developed the capacity to think critically about hard problems. To relegate our critical faculties to an activity you engage with during a small window of time each day similar to a muscle you exercise at the gym is asinine in my opinion. I firmly believe in the future, people like you will become a new underclass as they have willingly given up their ability to think.

[slibhb]

Again, people said the same thing about computers, about writing, and so on. Maybe it's true this time, but I think the presumption should be that it's not.

If people who use AI become an "underclass" then people will adapt and...not use AI. But that won't happen. People will use it to augment rather than replace, just like we use other, similar technologies.