Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
Microsoft AI

Microsoft Study Finds AI Makes Human Cognition 'Atrophied and Unprepared' (404media.co) 90

An anonymous reader shares a report: A new paper [PDF] from researchers at Microsoft and Carnegie Mellon University finds that as humans increasingly rely on generative AI in their work, they use less critical thinking, which can "result in the deterioration of cognitive faculties that ought to be preserved."

"[A] key irony of automation is that by mechanising routine tasks and leaving exception-handling to the human user, you deprive the user of the routine opportunities to practice their judgement and strengthen their cognitive musculature, leaving them atrophied and unprepared when the exceptions do arise," the researchers wrote.

Microsoft Study Finds AI Makes Human Cognition 'Atrophied and Unprepared'

Comments Filter:
    • by ewibble ( 1655195 ) on Monday February 10, 2025 @01:18PM (#65156335)

      Yeah it was obvious, it has obvious that use of things like google maps has reduced our ability to navigate. Yet we as a society continue on our path to becoming dumber. There is such a thing as making things too easy, too convenient. But while there is money to be made in it companies like Microsoft will continue to encourage it.

      • by sinij ( 911942 ) on Monday February 10, 2025 @01:25PM (#65156369)
        The danger is unconditionally trusting answers and not in having an ability to ask questions. The move from slide ruler to Matlab enabled us to do more, but fewer people able to do it well.
      • by Z00L00K ( 682162 ) on Monday February 10, 2025 @01:30PM (#65156393) Homepage Journal

        Personally I avoid AI for important stuff, I'm fearing that it's going to damage my intended writing if I let it re-write it with the consequence that it can become misinterpreted.

        Using AI to create an illustration is fine, even though that sometimes is wonky too.

      • by Drethon ( 1445051 ) on Monday February 10, 2025 @01:35PM (#65156407)

        Yeah it was obvious, it has obvious that use of things like google maps has reduced our ability to navigate. Yet we as a society continue on our path to becoming dumber. There is such a thing as making things too easy, too convenient. But while there is money to be made in it companies like Microsoft will continue to encourage it.

        I find ML makes things take longer. The LLMs come up with ideas I had not thought of, but wraps it in a ton of BS that does not produce the same result I was originally going for, so I spend a bunch of time identifying the good parts and bad parts, and making something that works out of what the LLM produced.

      • I was never much good at navigating, as I never had much directional sense even before google. But now google tells me how to get places accurately instead of some human doing it inaccurately, and once I've gone there a time or two then I remember how to get there without google. I don't think google moves the needle in one direction or another in this regard. Did paper maps decrease our ability to navigate?

        • by vux984 ( 928602 ) on Monday February 10, 2025 @03:41PM (#65156869)

          I don't think google moves the needle in one direction or another in this regard.

          Fascinating. I have a very different experience.

          If I use google maps to actually navigate (using turn by turn instructions), some part of my brain mostly shuts down and I WILL NOT learn the route quickly. Eventually, if its somewhere i go often, I'll more or less rote memorize it but i still won't really know where it is (e.g. in relation to other places).

          If make a point of looking at the map, the destination address street and street number, and mentally noting the turns, nearby crossroads, how many blocks up or down it is, and where it is relative to other sites I know, and then navigate there WITHOUT assistance. I do that a couple times, sometimes just once and I'll know how to get there, and where it is. And how to get from there to somewhere else.

          I know people who hop in the car state their destination, and then start following the instructions. I know of cases where the voice recognition fuck up and pick a similar named location in a different city or even country, and the person driving the car didn't pick up that anything was wrong until they were on a major highway headed into the mountains. And only then did the .. "hey where the fuck are we going?" kick in.

          Did paper maps decrease our ability to navigate?

          No. Because paper maps don't tell us where we are, where we want to go, or how to get there. We have to figure all that out for ourselves.

          Maps actually help me "crystallize" where things are in relation to each other, and improve my ability to remember where things are. Its very easy for me to be confused about where things are especially when the roads and rivers and train tracks meander a bit and cross each other at various places and angles. Seeing that on a map, once or twice, really helps a form an accurate mental picture of where things are.

          • I have little to no visual memory, so memorizing directions is not a thing I do. I have to learn where the roads go. I do have a spatial sense apparently, because I can still do that. Once I know where the roads are, I can find whatever, even though I can't picture it.

          • Nearly all navigation land animals do involves just following landmarks, which nowadays includes signs. You still do the same thing when navigating without google maps without necessarily having a good idea of where everything is in relation to one another. For the most part it doesn't even matter.

            You can pretty easily prove this to yourself. Go look at houses under construction. Notice how small the foundations appear in comparison to the completed model homes of identical size.

            When you're in the act of na

          • If make a point of looking at the map, the destination address street and street number, and mentally noting the turns, nearby crossroads, how many blocks up or down it is, and where it is relative to other sites I know, and then navigate there WITHOUT assistance. I do that a couple times, sometimes just once and I'll know how to get there, and where it is. And how to get from there to somewhere else.

            This aligns perfectly with my own experiences and my observations of others trying to get around foreign places.

      • Yeah it was obvious, it has obvious that use of things like google maps has reduced our ability to navigate. Yet we as a society continue on our path to becoming dumber. There is such a thing as making things too easy, too convenient. But while there is money to be made in it companies like Microsoft will continue to encourage it.

        And spell checkers have significantly decreased the ability to spell. Yet, it might be arguable that spell checkers enable trading off a low-level skill like spelling to allow humans more time and resources for high-level skills like composition. Compilers decimated the ability to code in assembly. All successful tools repeat this takeover of low-level skills to enable more efficient high-level skills. This has happened over and over again, and we generally laud this progress (well, at least after it ha

        • spell checkers enable trading off a low-level skill like spelling to allow humans more time and resources for high-level skills like composition

          You seriously believe that people put more effort into their thoughts and compose better, with spell checkers? Have you evidenced this, at all? I'd argue the opposite happens. You get a faster lane between thought and expression, with minimal time to think, process, filter, improve. Which are things that improve composition. And the time relieving the brain of the extra effort is not used to train the brain in other skills, it's just brain staying lazy ("resource efficient") and just atrophies. Agreed abou

        • by narcc ( 412956 )

          Yet, it might be arguable that spell checkers enable trading off a low-level skill like spelling to allow humans more time and resources for high-level skills like composition.

          Maybe back when spell checking was a separate process. These days, the little red line appears immediately, calling attention to your mistake and interrupting your train of thought. It's a distraction that slows you down. That's at its best.

          Worse, it's constantly there, lurking in the background, just waiting to interrupt you. Your attention is constantly divided. You do your best avoid setting it off so it won't distract you, but that just distracts you more. 'i' before 'e' except after 'c' ... someth

        • The technical school I went to demanded written reports whenever a class/teacher asked to create one. Spelling, grammar, penmanship, amount of words....all items that removed half or a full point off the total score with every error. It was the Netherlands, so test scores were between 1 and 10. No forgiveness could be expected from the teachers. And quite often you had to make 3 to 4 multipage reports per week, next to 3 to 4 hours of homework every day. And school started often at 08:30 and ended at 17:30.

      • by Torodung ( 31985 )

        That's why folks should use Google Maps to observe, and then turn it off and see how they do on their own for common trips. Or use the map without a stated route to get your bearings. Google hates this and will offer routes, but you can ignore it and just use the map.

        Additionally, I frequently refuse turns that Maps recommends because of inclement weather or just knowing a better route. Maps has a predilection to send you down back roads in a snowstorm, which is a bad idea. It has no option for "I'm in a bl

      • The Amish people agree with you.

    • It's really obvious. Just ask anyone that moved from tech to management and you'll find skills atrophied. AI is even worse, as at least a tech manager might still review their subordinates work, whereas AI mostly gives you what you ask for. Let's be honest - you run the code and if it works you include it! New starters are using AI and missing the fun of spending two days on a single problem - that is what gives you life skills
  • by _dj6_ ( 8250908 ) on Monday February 10, 2025 @01:06PM (#65156305)
    Our overlords absolutely love this aspect of AI. It can dumb down the population at a rate they didn't dare dream of before! ;)
  • by sinij ( 911942 ) on Monday February 10, 2025 @01:09PM (#65156315)
    For some reason ancient Greeks understood the danger of asking Oracles, but we do not understand similar dangers of asking AI.
  • Use AI to maximize productivity &out compete your peers: Get a job!

    Use AI and lose your ability to handle irregular situations: Lose a job!

    Everything will be fine as long as nothing goes wrong in a novel way.

  • by ebunga ( 95613 ) on Monday February 10, 2025 @01:18PM (#65156337)

    I mean, sure, the CEOs keep trying to put a positive spin on it, but every single study has shown AI as it is right now to be a big negative. Maybe we should stop making the world a worse place?

    • by HiThere ( 15173 )

      Every single honest study of ANYTHING show that it has negative effects. What we can't do is agree on what the balance is. We can't even agree on coffee vs. tea. (Or vi vs. emacs.)

    • I mean, sure, the CEOs keep trying to put a positive spin on it, but every single study has shown AI as it is right now to be a big negative. Maybe we should stop making the world a worse place?

      Making the world a worse place can be profitable. And profit is good. So, really, we're making things worse for the good of the world! Or something. I dunno. My brain has atrophied from a steady diet of Marvel movies and automation posing as AI.

  • by Junta ( 36770 ) on Monday February 10, 2025 @01:20PM (#65156347)

    Those who are able to or believe they are able to use LLMs to do their jobs may not have the same cognitive load as those who have failed to find LLM approaches up to the task of doing all their work?

    There's an executive I know that has gone hard in on LLM. Thing was he never impressed me before, and his switch to using LLM to generate his correspondence and digest meetings for him hasn't really made him any worse than he already was. He is really bad at his job now, but he was before too.

    • It's probably causation, if our many models of human cognition, neural connectivity, brain plasticity, and clinical cerebral atrophy are to be believed.

      There are also a lot of lazy idiots that are attracted to an easy way out, so are likely well represented in the LLM user base.

    • What do you mean or? Correlation is strictly necessary in any causal relationship.
  • by wakeboarder ( 2695839 ) on Monday February 10, 2025 @01:30PM (#65156391)

    If you use it as a time saver for doing repetitive tasks it's great. If you are using it to do your job, then you aren't going to think enough. Also if you are letting AI do your job for you, you may want to get a new job, it's likely that AI might be coming for your job.

  • by nimbusarch ( 10483592 ) on Monday February 10, 2025 @01:36PM (#65156409)
    I also to some degree blame modern web design/publishing and the ad driven market as to why its so appealing. The desire to keep people on a page or clicking through sub pages for ad revenue while fighting hostile ad mechanics has added to the desire to use a tool that gets to the damn point. My patience for fighting to extract direct information to simple questions has faded a lot and AI has made that easier. I ask a question, I get an answer. No need to dig through someone's narrative word salad (also often AI written). Probably something I need to work on.
  • Shop Floor (Score:5, Interesting)

    by bill_mcgonigle ( 4333 ) * on Monday February 10, 2025 @01:40PM (#65156427) Homepage Journal

    The old Mech-E's would say, "90% of innovation happens on the shop floor."

    Meaning that when designers and industrial engineers throw their design over the wall, that's the first ten percent.

    Only through the iteration of the manufacturing process does reality come into focus.

    When American beancounters decided to move all their manufacturing to China but keep their "core competencies" in the US, their fatal conceit was ignoring this. As a consequence China is starting to pull ahead in nearly all areas.

    The same lesson will apply to AI assistants. If a human is only involved in the superficial aspects they will ultimately fail to understand the problems.

    A flip side is that perhaps people of lower IQ will be able to do more. That is good.

    But the flip-flip side is that we may end up with a scenario where the highest IQ people are expected to work the hardest and everybody else will be able to work easily or not at all. That is inherently unstable.

    We need to consider risks at third-orders and beyond. That's not (yet) something you can ask an AI to do.

    • I wish I had mod points to mod this up even more.

      If a human is only involved in the superficial aspects they will ultimately fail to understand the problems.

      THAT is the reality, instead of the fantasy of "let the AI do all this menial tedious low-level thinking, so that we are free to do the high-level brainy stuff".

  • by n0w0rries ( 832057 ) on Monday February 10, 2025 @01:50PM (#65156463)

    When you are too stupid to think for yourself, you will no longer be useful to AI, because you will no longer be a source of learning.

    I was watching some youtube channels where these guys make stuff with 3d printers and all kinds of high tech tools and 3d design. Pretty amazing. Then they pitch to their audience to use these apps to project manage your build, which that data is basically training AI how to design and build things.

    Things are going to accelerate very rapidly.

  • Election Results? (Score:1, Flamebait)

    by tsqr ( 808554 )

    Could this explain the results of the last Presidential election?

    Disclaimer: I'm registered non-affiliated, and I live in a state that hasn't been carried by a Republican since Reagan. I didn't vote for Trump or Harris My state was won by Kamala by a very comfortable margin.

  • No, I don't believe that AI has ever had a single conscious though.
    • by narcc ( 412956 )

      No, I don't believe that AI has ever had a single conscious though.

      Why anyone thinks otherwise is the real mystery...

  • AI Makes Human Cognition 'Atrophied and Unprepared'

    Speed dial makes people forget phone numbers.

  • by az-saguaro ( 1231754 ) on Monday February 10, 2025 @02:48PM (#65156679)

    I read the original article, pdf at https://www.microsoft.com/en-u... [microsoft.com]
    Quite a few things are odd or interesting.
    One is selection bias.

    They surveyed 319 "knowledge workers" which they identified from a survey where people self reported if hey used AI, from which they recruited those who used it every week.
    To me, that says these people are either (1) legitimately curious intellectual people who wanted to see what the fuss is about, tried AI tools, found something genuinely compelling, and continue to use it as early adopters of the good side of a new technology; or (2) lazy short-cut seeking intellectual lightweights who see AI as a cheap-and-easy run around doing the necessary real work. [Yes, I recognize that there is a spectrum and exceptions, and shades of gray between.]

    Then, the participants were asked to give three examples of how they use AI, covering three usage cases, BUT (if I understand their methods correctly), if they didn't have examples of one or two categories, then they could use hypothetical examples (paraphrasing).

    Then, they self reported their perceived level of critical thinking versus passive use of the AI information. (The paper is 20 pages of lots of words, but not always clearly expressed explained - makes me wonder if they did or didn't use AI to help obfuscate the whole thing.) So, the same people using AI for perhaps pointless or silly tasks that do not need AI (some of the use examples were rather basic grammar school level tasks) are then explaining their perceived level of intellectual engagement, then the AI-sponsoring research team is analyzing those answers with "statistics". It simply isn't the way good or objective studies are usually designed.

    Despite that, the conclusions ring true - AI makes people dumber, or restricts their "brain exercises" to limited opportunities.
    But I am not sure how you generalize that to the whole population or workforce. What is scary to me is that these "knowledge workers" who reportedly use AI for mundane tasks that don't need AI implies they are not critical thinkers in the first place. But they are in responsible workplace positions, the types that become middle management, implying that that already broken and irrelevant tier of organizational structure will only get stupider with time as they get ever-better tools to hide their incompetence and fundamental lack of intellectual, technical, or entrepreneurial skills.

    I know, I am extrapolating a lot from just a little, but this study is weird, but still somehow correct and prophetic.

  • Ok so in the interest of public service they are going to stop trying to shove GenAI driven automation features into every aspect of Windows and Office right? right?

  • There has been a decline in critical thinking and using analytical ability over the past 40 years, possibly longer. Humans in general do not look at things and question, they take what they are told, and only if things seem obviously questionable do they look into it. The idea of questioning EVERYTHING, even stuff you agree with is alien to most people. Ask people to come up with solutions to problems, and you get a bunch of people who believe that you need a specialist to look into it, because they j

  • "leaving them atrophied and unprepared when the exceptions do arise" - sounds like Microsoft standard operating procedure all along any way. All they can do now is shift blame to the AI ... which come to think of it may have been the goal all along. hmmm..

    • "leaving them atrophied and unprepared when the exceptions do arise" - sounds like Microsoft standard operating procedure all along any way. All they can do now is shift blame to the AI ... which come to think of it may have been the goal all along. hmmm..

      Microsoft made me worse at computing, without a doubt. Setting up networks in Win95 and Win98 always seemed to involve trying the exact same thing at least three times, and each time there would be a different result. Not a good way to learn cause and effect.

  • Because last time i remember there were concerns about mental health, regarding how much we have to cognitively peocess unnaturally to our species and how it makes us burnt out
  • most important word: self-reported. for what we know, their actual critical thinking skills may even have got better! or not changed at all!
    • How do you improve YOUR critical thinking, when you get all your answers from the same source machine? (that is known to be occasionally inaccurate).

      "Critical thinking is the process of analyzing available facts, evidence, observations, and arguments to make sound conclusions or informed choices"

      When our facts, evidence, observations and arguments come from the same source, we are just being parrots, parroting the hallucination of critical thinking of the machine/source. It's not only critical thinki
  • by Petersko ( 564140 ) on Monday February 10, 2025 @04:31PM (#65157043)

    This has been happening for a long time, but it really got cooking when the internet got a foothold.

    My father was a heavy duty mechanic by trade. He also built houses, cut hair, decorated wedding cakes for extended family, hunted, fished, fixed appliances... whatever needed doing, he figured out how to do it. I'm 54, and to a lesser extent, I'm similar. I can do everything on the list except I'm not a mechanic, save for minor motorcycle repairs, and my house building skills are limited. I can frame, roof, drywall, paint, insulate, wire... but I couldn't build a deck to code without tons of research.

    The thing is, disconnect us from the internet and drop us in a problem, and we'll probably do alright. Well... he disconnected completely in 2022... but the point remains.

    Having access to the world's knowledge makes you a conduit for it, but nothing more. You understand little if anything of it. Of course your function atrophies. You don't retain knowledge you didn't earn. If "google the error code" is the termination of your investigation because the internet is down... that's a shame.

    "A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects." - Robert Heinlein (Lazarus Long)

    • Agree completely.
      Well stated.

      You sell yourself short though.
      " ... but I couldn't build a deck to code without tons of research."
      You can frame, so you can build a deck. Building to code - they vary place to place, so no matter what, you would have to do some research to know the code. That is not a deficiency of skills, just the realities of the legitimate real world. Many times, "knowledge" is a set of skills adaptable to many situations with a bit of incidental lookup. Nobody, unless a full time profes

    • Wasn't "code" much simpler when your father built according to it?

  • ... practice their judgement ...

    Have you seen the US government? Forget all the bad actors: I know there are hundreds of them, forget them. The biggest bad actor is the voters, who found the answers to their problems in a 10-second sound-bite. I know the system is rigged against US voters but their refusal to think "What happens next?" is a lack of judgement and even a lack of adult behaviour. That lack of responsibility was driven by a demand for intellectual simplicity in general, not for AI slop, specifically.

    Many of Trump's poli

    • by Gilmoure ( 18428 )

      As a kid who expected to die from a nuclear exchange before I graduated high school (class of '85), I'm kinda amazed we made it this far but not really suprised to see us spinning the drain now.

      Part of me thinks it would be a good thing to get one of those 'How To Rebuild Civilization' books and store it away in a hidden location and the rest of me is like, maybe find such a book but for raccoons. I have a feeling they'll be the B-Team that evolves into something interesting and destructive in North America

  • Mah floaty chair has crashed and spilled my BnL drink!

  • By the very design, AI can't think, just pattern match. I use it daily between IntelliJ, and CoPilot and I honestly think I spend more time fixing AI mistakes than they save me if I just turn off the plugins. I asked it to write a simple Java unit test today, it generated code that LOOKED like it worked, but it didn't even compile. Fortunately the compiler caught it, but TBH, it would have taken me a minute or 2 to figure out it was completely wrong without the compiler....it also was a lot more complex
    • by narcc ( 412956 )

      I honestly think I spend more time fixing AI mistakes than they save me

      That's the conclusion most people reach once the novelty wears off and they stop blaming themselves for the unrealized gains. I've been saying for years now that those early subjective reports of increased productivity could be explained completely by those users being more focused on work because they were having fun playing with a new toy.

      • I honestly think I spend more time fixing AI mistakes than they save me

        That's the conclusion most people reach once the novelty wears off and they stop blaming themselves for the unrealized gains. I've been saying for years now that those early subjective reports of increased productivity could be explained completely by those users being more focused on work because they were having fun playing with a new toy.

        AI is the new religion and at my employer, management wants to know we're using the latest tools and toys. I actually got mildly scolded when they figured out I didn't activate CoPilot in my IDE. I am also getting scolded by another manager because my pull requests are too concise and tight. In other words, by quality alone, they can tell it's written by a very experienced and enthusiastic human...not robotic slop generated by an AI. Developers at my company are supposed to be using AI as much as possib

  • Does anyone remember when people had very serious debates about how calculators were making the human race stupider? Anyone? Bueller?

    We have this same exact (dumb) discussion anytime we invent a new tool that helps save us brain cycles, and it always sounds like everyone involved is the preacher's wife on the Simpsons ("won't someone please think of the children?!?"). In other words, they don't have rational concerns, just irrational fear.

    Sure there will be some growing pains, as there is with any new te

    • by narcc ( 412956 )

      Does anyone remember when people had very serious debates about how calculators were making the human race stupider? Anyone?

      There is still serious debate. I tend to fall on the side of the naysayers, given how often I've seen kids reach for a calculator to multiply a number by a power of 10.

      Does anyone remember when people had very serious debates about whether children should memorize multiplication tables? Anyone?

      It turns out that the kids that did have a massive advantage over the kids that didn't. Jo Boaler, you should be ashamed of yourself. More than just saving time with trivial calculations, the real saving start when

      • The mental exercise of doing math manually has benefits and while it may literally be as pointless as lifting weights or jogging the EXERCISES produce benefits that are now the point of doing it while initially it was just to perform the manual operation.

        The understanding of what is actually going on can be emphasized or realized by doing the fundamentals yourself. This even happens in programming where it can take less time to re-code something than it does to study the existing code; or compare the two a

      • > Does anyone remember when people had very serious debates about whether children should memorize multiplication tables? Anyone?

        Yes; I actually went through school.

        We had to memorise all up to the 12 times table.

        I was really chuffed when I started figuring out tricks and methods to calculate the answer vs memorise it, I still use that till this day, especially with 9's and 8's.

    • > Does anyone remember when people had very serious debates about how calculators were making the human race stupider

      Yes; I went through school.

      And I agreed. Even if I hated to.

      I got to the point where I couldnt understand the concept of a prime number without a calculatir.

      Also dont ask me to divide an odd number, ever. Only even numbers exist without a calculator :D

    • Not even remotely in the same boat.

      Our biggest struggle with newer talent is a fundamental inability to problem solve. When faced with a problem they will either play all manner of avoidance cards or they will reach for their favourite LLM and all the garbage it scraped from reddit/github/etc.

      Dealing with problems is one of humanity's most enduring and important skills and training multiple generations to offload this on LLMs is most assuredly not going to produce good outcomes.

  • ...can now only get aroused around those with an extra set of "fun parts". Expect birth rates to decline further.

    Although microplastics may eventually create real people who match that. Is futurism self-patching?

    Schrodinger's Anthropic Principle, or God has a strange sense of humor?

  • As humans continue to develop our now mostly metaphorical but maybe someday literal exocortices they will grow new pathways, new weights. Yes some capacity will be pruned, I can't remember as much Hellenic poetry as, for example, Homer (ÎOEμÎÏÎÏ), but I have a lot of memories he doesn't... like how to operate a car, ps5, navigate the web, find my way around Skyrim, or have useful conversations with AI agents. Someday hopefully I will have a BCI or artificial cortex I can
  • Politzania, brave strong and true
    Politzania, we all love you
    We'll smite our foes for we are right
    And God is on our side

    Politzania, red, white and green
    Politzania, reigning supreme
    Victors in war
    Champions of peace
    Unto eternity

    Everybody now
    La la la la la la la la la la la la
    La la la la la la la la la la la la la la
    La la la la
    La la la la la la
    We're the masters of the world

  • This isn’t new, it has been happening for at least a decade.

    It's been happening since we started reaching for phones in our pockets to "remember" things by googling. Back then I remember articles about the changes happening in peoples (especially kids) brains that durned people into efficient researchers, able to find practically anything on the net and quickly. Their brains were geared towards forming search queries, and navigating paths to find information. Sounds neat, but it is at the expense of

  • It seems like people who already have strong critical thinking skills and creative patterns of thought can benefit enormously. If you are a person like me who has more ideas than you can even begin to explore at any one time, much less implement, AI tools can allow you to chase WAY more of your original thinking. I absolutely see the crowd that might already be a bit slow-thinking or intellectually lazy moaning disproportionately about how useless AI is. My take is that they just aren't smart or creative

MESSAGE ACKNOWLEDGED -- The Pershing II missiles have been launched.

Working...