Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

publichealthpolicyjournal.com

Post reply on HN
Print view View original

251–260 of 589 posts

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#251

Post by jbstack »

mzajc wrote: post #175
jbstack wrote: post #138

Earlier quoted context omitted.

> I've experienced cognitive decline when using AI tools a bit too much to help solve programming-related problems. When dealing with an unfamiliar programming ecosystem it feels so easy and magical to just keep copy / pasting error outputs until the problem is resolved. Previously solving the problem would've taken me longer but I would've also learned a lot more. I've had the opposite experience, but my approach is…

> I ask follow up questions to make sure I understand why the AI's answer works. I've tried a similar approach and found it very prone to hallucination[0]. I tend to google things first and ask a LLM as fallback, so maybe it's not a fair comparison, but what do I need a LLM for if a search engine can answer my question. [0]: Just the other day I asked ChatGPT what a colonn (':') after systemd's ExecStart= means. The…

It's a tradeoff. After using ChatGPT for a while you develop somewhat of an instinct for when it might be hallucinating, especially when you start probing it for the "why" part and you get a feel for whether its explanations make sense. Having at least some domain knowledge helps too - you're more at risk of being fooled by hallucinations if you are trying to get it to do something you know nothing about.

While not foolproof, when you combine this with some basic fact-checking (e.g. quickly skim read a command's man page to make sure the explanation for each flag sounds right, or read the relevant paragraph from the manual) plus the fact that you see in practice whether the proposed solution fixes the problem, you can reach a reasonably high level of accuracy most of the time.

Even with the risk of hallucinations it's still a great time saver because you short-circuit the process of needing to work out which command is useful and reading the whole of the man page / manual until you understand which component parts do the job you want. It's not perfect but neither is Googling - that can lead to incorrect answers too.

To give an example of my own, the other day I was building a custom Incus virtual machine image from scratch from an ISO. I wanted to be able to provision it with cloud-init (which comes configured by default in cloud-enabled stock Incus images). For some reason, even with cloud-init installed in the guest, the host's provisioning was being ignored. This is a rather obscure problem for which Googling was of little use because hardly anyone makes cloud-init enabled images from ISOs in Incus (or if they do, they don't write about it on the internet).

At this point I could have done one of two things: (a) spend hours or days learning all about how cloud-init works and how Incus interacts with it until I eventually reached the point where I understood what the problem was; or (b) ask ChatGPT. I opted for the latter and quickly figured out the solution and why it worked, thus saving myself a bunch of pointless work.

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#252

Post by card_zero »

joenot443 wrote: post #241
card_zero wrote: post #234

Earlier quoted context omitted.

I can fart in the general direction of the code and that's a kind of abstraction too. It distills my intent down to a simple raspberry noise which could then be interpreted by sufficiently intelligent software.

Sort of like when people argue about "what is art?", some folks find it clever to come up with the most bizarre of examples to point to, as if the entire concept rests on the existence of its members at the fringe. Personally, I think if your farts are an abstraction that you can derive useful meaning from the mapping, who are we to tell you no?

So, can derive useful meaning from is the point of contention.

(Also: bizarre examples = informative edge cases. Sometimes.)

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#253

Post by nancyminusone »

sudosteph wrote: post #88

Meanwhile my main use cases for AI outside of work: - Learning how to solder - Learning how to use a multimeter - Learning to build basic circuits on breadboxes - learning about solar panels, mppt, battery management system, and different variations of li-on batteries - learning about LoRa band / meshtastic / how to build my own antenna And every single one of these things I've learned I've also applied practically t…

As someone who does these things, I am curious to know how and why you would choose AI.

Working these from text seems to be the hardest way I could think to learn them. I've yet to encounter a written description as to what it feels like to solder, what a good/bad job actually looks like, etc. A well shot video is much better at showing you what you need to do (although finding one is getting more and more difficult)

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#254

Post by bckr »

e3bc54b2 wrote: post #228
CuriouslyC wrote: post #189

Earlier quoted context omitted.

We're just moving up the abstraction ladder, like we did with compilers. I don't care about the individual lines of code, I care about architecture, code structure, rigorous automated e2e tests, contracts with comprehensive validation, etc. Rather than waste a bunch of time pouring over agent PRs I just make them jump over extremely high static/analytic hurdles that guarantee functionality, then my only job is to ide…

As the other comment said, LLMs are not an abstraction. An abstraction is a deterministic, pure function, than when given A always returns B. This allows the consumer to rely on the abstraction. This reliance frees up the consumer from having to implement the A->B, thus allowing it to move up the ladder. LLMs, by their very nature are probabilistic. Probabilistic is NOT deterministic. Which means the consumer is neve…

The LLM is not part of the application.

The LLM expands the text of your design into a full application.

The commenter you’re responding to is clear that they are checking the outputs.

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#255

Post by falcor84 »

lexandstuff wrote: post #146
falcor84 wrote: post #93

Earlier quoted context omitted.

What does that mean? If you're still paying for a Claude Code, you are supposedly getting increased productivity, right? Or otherwise, why are you still using it?

I find it useful. A nice little tool in the toolkit: saves a bunch of typing, helps to over come inertia, helps me find things in unfamiliar parts of the codebase, amongst other things. But for it to be useful, you have to already know what you're doing. You need to tell it where to look. Review what it does carefully. Also, sometimes I find particular hairy bits of code need to be written completely by hand, so I ca…

I don't think the concern is that non-experts would manage large software systems, but that experts would use it to manage larger software systems on their own before needing to hire additional devs, and in that way reduce the number of available roles. I.e. it increases the "pain threshold" before I would say to myself "it's worth the hassle to hire and onboard another dev to help with this".

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#256

Post by Jensson »

elAhmo wrote: post #231
aprilthird2021 wrote: post #204

Earlier quoted context omitted.

> We're just moving up the abstraction ladder, like we did with compilers. We're not because you have to still check every outputted code. You didn't have to check every compilation step of a compiler. It was testable actual code, not non-deterministic output from English language input

I would bet a significant amount of money that many LLM users don’t check the output. And as tools improve, this will only increase. The number of users actually checking the output of a compiler is nonexistent. You just trust it. LLMs are moving that direction, whether we like it or not

> The number of users actually checking the output of a compiler is nonexistent. You just trust it.

Quite a few who work on low level systems do this. I have done this a few times to debug build issues: this one time a single file suddenly made compile times go up by orders of magnitude, the compiler inlined a big sort procedure in an unrolled loop, so it added the sorting code hundreds of times over in a single function and created a gigantic binary that took ages to compile since it tried to optimize that giant function.

That is slow both in runtime and compile time, so I added a tag to not inline the sort there, and all the issues disappeared. The sort didn't have a tag to inline it, so the compiler just made an error here, it shouldn't have inlined such a large function in an unrolled loop.

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#257

Post by chii »

CuriouslyC wrote: post #189
benterix wrote: post #173

Earlier quoted context omitted.

We are literally witnessing the skills split right in front of our eyes: (1) people who are able to understand the concepts deeply, build a mental model of it and implement them in code at any level, and (2) people who outsource it to a machine and slowly, slowly lose that capability. For now the difference between these two populations is not that pronounced yet but give it a couple of years.

We're just moving up the abstraction ladder, like we did with compilers. I don't care about the individual lines of code, I care about architecture, code structure, rigorous automated e2e tests, contracts with comprehensive validation, etc. Rather than waste a bunch of time pouring over agent PRs I just make them jump over extremely high static/analytic hurdles that guarantee functionality, then my only job is to ide…

> my only job is to identify places where the current spec and the intended functionality differ and create a new spec to mitigate.

and to be able to do this efficiently or even "correctly", you'd need to have had mountains of experience evaluating an implementation, and be able to imagine the consequences of that implementation against the desired outcome.

Doing this requires experience that would get eroded by the use of an LLM. It's very similar to higher level maths (stuff like calculus) being much more difficult if you had poor arithmetic/algebra skills.

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#258

Post by daveguy »

e3bc54b2 wrote: post #228
CuriouslyC wrote: post #189

Earlier quoted context omitted.

We're just moving up the abstraction ladder, like we did with compilers. I don't care about the individual lines of code, I care about architecture, code structure, rigorous automated e2e tests, contracts with comprehensive validation, etc. Rather than waste a bunch of time pouring over agent PRs I just make them jump over extremely high static/analytic hurdles that guarantee functionality, then my only job is to ide…

As the other comment said, LLMs are not an abstraction. An abstraction is a deterministic, pure function, than when given A always returns B. This allows the consumer to rely on the abstraction. This reliance frees up the consumer from having to implement the A->B, thus allowing it to move up the ladder. LLMs, by their very nature are probabilistic. Probabilistic is NOT deterministic. Which means the consumer is neve…

> An abstraction is a deterministic, pure function, than when given A always returns B.

That is just not correct. There is no rule that says an abstraction is strictly functional or deterministic.

In fact, the original abstraction was likely language, which is clearly neither.

The cleanest and easiest abstractions to deal with have those properties, but they are not required.

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#259

Post by __loam »

elAhmo wrote: post #231
aprilthird2021 wrote: post #204

Earlier quoted context omitted.

> We're just moving up the abstraction ladder, like we did with compilers. We're not because you have to still check every outputted code. You didn't have to check every compilation step of a compiler. It was testable actual code, not non-deterministic output from English language input

I would bet a significant amount of money that many LLM users don’t check the output. And as tools improve, this will only increase. The number of users actually checking the output of a compiler is nonexistent. You just trust it. LLMs are moving that direction, whether we like it or not

Well they're not improving that much anymore. That's why Sam Altman is out there saying it's a bubble.

Re: MIT Study Finds AI Use Reprograms the Brain, Leading to Cognitive Decline

#260

Post by MarkusWandel »

Muscles atrophy from lack of use - as an aging cyclist with increasing numbers of e-bikes all around, I think I may some day have to use one because of age, but what are all these younger people doing, cheating themselves out of exercise?

And so it is with many things. I wrote cursive right through the end of my high school years, but while I can type well on a computer, I have trouble even writing block lettering without mistakes now, and cursive is a lost cause.

Ubiquitous electronic calculators have eroded the heroic mental calculation skills of old. And now artificial "thinking machines" to do the thinking for you cause your brain to atrophy. Colour me surprised. The Whispering Earring story was mentioned here just recently but is totally topical.

https://croissanthology.com/earring

Post reply on HN