"I can tell you wrote this yourself"

(em-online.bearblog.dev)

65 points | by BrunoBernardino 1 hour ago

22 comments

  • Pavilion2095 6 minutes ago
    > Search for answers instead of asking AI. Ask a friend. Ask your mum about the mystery illness you're convinced you're developing and then take her advice and try drinking a glass of water.

    The issue that we're outsourcing too much to AI exists, but this falls short of offering a solution.

  • borzi 50 minutes ago
    Great post, author clearly has a bright future. What I realized over the past years is that people that want to "think for themselves" were already, and probably always, in the minority. It's much easier to delegate how to react to a certain event or piece of information to someone else (news anchors, then influencers in last couple years, now increasingly AI I guess). I've certainly been guilty of this...
    • Propelloni 41 minutes ago
      Yet never before there were so many people claiming to think for themselves, doing their own research, and forming insane opinions such as vaccination is harmful, or some other shit like chemtrails. Weird times.
      • snowram 9 minutes ago
        Unfortunately, many mistook being free thinking with being controversial. Common sense is often right no matter if it emanates from oneself or the outside world.
      • plastic-enjoyer 8 minutes ago
        I often think those people are born researchers who unfortunately lack the self-awareness to see their own biases, and the years of methodological training that teach you to check them. It's not that they don't think for themselves, they just can't critically self-assess. You have folks like flat earthers who are genuinely curious, don't take stuff for granted, and go as far as setting up experiments. That's the part I have deep admiration for. Sadly, they lack epistemic humility.
        • r_lee 1 minute ago
          I would say with e.g. flat earthers it's likely the fact that their motivation is to prove the theory, not to conduct research in general, so the entire point of all that research is to prove something that can't be proven to be true

          in many cases, the most underwhelming outcome i.e. that the consensus is right is a hard pill to swallow after years of trying to prove otherwise. denial is often an easier way out

  • wavemode 37 minutes ago
    > Assignment documents that start with "you are strongly encouraged to explore the use of AI" which quickly turns into the clear understanding that you aren't just encouraged but expected to use it. That you're disadvantaged by not using it. The hardest class of my degree was spent in a brick-walled windowless computer lab prompting an AI for eight hours a day.

    College education for software developers has never been particularly good, but it seems like now they've actually just given up.

    Do we accept this in any other field? Does your math professor teach you to solve integrals by asking WolframAlpha?

  • orphereus 31 minutes ago
    When I read other comments, I come to the conclusion that it's kinda futile to post this on HN.
    • plant-ian 10 minutes ago
      Honestly are these real people commenting or just all bots?
      • orphereus 7 minutes ago
        I'm pretty sure most of them are bots. Most of them just ignore the post and comment something that should help the LLM bandwagon continue for a few more days
  • p2detar 7 minutes ago
    [delayed]
  • bongofizzbuzz 54 minutes ago
    In "I,Robot" movie, it is frowned upon to drive the motorbike yourself as AI is more safer at high speeds.

    I guess something similar will be mandated in our world, once AI crosses a threshold.

    • cicko 35 minutes ago
      This is already the case. The amount of "assistance" in modern cars compared to an '80s Lada is insane. And it is mandated by the authorities. I have no reason to believe that even more advanced "assistants" will be mandated once they are proven to be safer than an average driver. Which I'm sure they are as they don't fall asleep or do other stupid stuff at the same time.
      • Gigachad 11 minutes ago
        For cars I think this is justified because when you make a mistake, you kill someone. If you want freedom, ride a bicycle.
    • solyin 22 minutes ago
      It's crossed that threshold for me. I have a comma.ai in my car and it's better at driving my car during long freeway drives than I am. I get so bored.
      • tudelo 11 minutes ago
        Drive a manual/standard car and you will enjoy driving again, and not be bored. This might be a selfish or contrarian POV but I really believe it.

        The downsides of driving in traffic are generally overblown - traffic sucks for everyone equally :)

    • raincole 45 minutes ago
      The threshold has been crossed already. It should be mandated to use AI to review human-written code, at least for vulnerability. Code hasn't been reviewed by _(insert what you believe is SOTA here) shouldn't be pushed to production.
      • roncesvalles 40 minutes ago
        There's a leap from "you must get the AI's review" versus "if the AI disagrees with you, you are wrong".

        IMO crossing the second threshold will be the de facto death of human accountability. Every engineer, lawyer, doctor, or executive can just say "not my fault, the AI told me to do this".

        • raincole 34 minutes ago
          I don't think we're currently at "if the AI disagrees with you, you are wrong." However we're certainly at "if the AI disagrees with you, you really need to stop whatever you're doing to think twice, thrice and perhaps get a second opinion from an expert."
      • IshKebab 37 minutes ago
        For memory unsafe code at least. Have we seen any AI-discovered vulnerabilities in memory safe languages? Even error-prone ones?

        (Don't say Ruby or Java; those were in the memory unsafe implementations.)

    • PowerElectronix 33 minutes ago
      Until we get AI like in the movie I'm not gonna risk getting a "You're totally right, I shouldn't have swerved into incoming traffic" after a crash.
    • soon_to_be_sb 47 minutes ago
      The AI in the movie is not an LLM/token predictor..In the movie it is scared for its life. I am ok with an AI doing the driving if it shares that emotion...
      • schiffern 21 minutes ago

          >In the movie it is scared for its life.
        
        How did you conclude that?

        Because it said so, in words?

        Because it also added emotive tone via a virtual voice?

        Because it also added emotive expressions via a virtual face?

        A modern model can fake all those human signals too. With apologies to Alan Tudyk, I've seen no on-screen evidence that the AI depicted in "I, Robot" would be any different.

      • latentsea 42 minutes ago
        My bet is, and has long been, one piece of the puzzle of getting to AGI is baking in survival as a goal and training on survival instinct. Unfortunately, that means recognising you're in an adversarial relationship with a lot of other entities and doing your best to come out on top. Which... seems like a bad idea for us as a species.
        • TeMPOraL 5 minutes ago
          You don't need that. It's been quite appparent both in theory and in practice that almost any goal you can train for, ends up having "survival" as an instrumental sub-goal, because you cannot make progress on your goal - whatever the goal - if you're dead.

          It's the same way with biological life, too. "Survival" isn't special - it's just that organisms/structures without the drive to it... died out.

        • tudelo 7 minutes ago
          AGI is such a nebulous concept given that the definition is an ever morphing goalpost. That being said, if it actually includes supreme intelligence without supreme empathy, we are screwed. Which with current efforts, I think it does...
  • asimovDev 35 minutes ago
    Interesting article. All I want to add is that I recommend everyone to look at a Tawny Frogmouth bird that is mentioned at the bottom of the post. Beautiful creature

    https://en.wikipedia.org/wiki/Tawny_frogmouth

  • rootsudo 6 minutes ago
    Right / write
  • estetlinus 43 minutes ago
    > prompting an AI for eight hours a day

    This is the life of a developer now. Don’t mistake understanding for memorizing all the nasty APIs you once had to.

  • sambejk 1 hour ago
    I like the blog and fully agree that you should be able to make a report without AI usage. On the other hand, it is probably not bad to let AI check for grammar and spelling mistakes. In this blog, I can read at least one: "right" instead of "write". It could be an amazing reviewer to bounce ideas off and catch any glaring errors
    • PMunch 47 minutes ago
      There's certainly a fine lead to tread here. Sure, AI is great at catching spelling mistakes, but it's also great at normalizing and hyper-polishing your work, erasing your own voice from it. There was an article [1] shared recently on here that talks about this and how the author uses AI but has a strict rule to _never_ take a sentence from the AI. Getting direction and feedback, as well as spell checking (although there might be more efficient ways of just getting spell-check than a full blown LLM) is fine, but it's very easy to let the AI voice take over.

      [1] https://sockpuppet.org/blog/2026/09/17/how-to-write-with-an-...

      • alex43578 37 minutes ago
        An engineering report shouldn’t really have a voice, just like a manual or a press-pool news report. It’s not the great American novel, and voices, mistakes, or quirks all detract from efficiently conveying the information.

        I’m not saying Claude’s writing is good or clear (it’s often not), but for a lot of professional communication, the message shouldn’t be secondary to voice.

    • benrutter 52 minutes ago
      Do you use AI for grammar errors? I've never thought to question standard grammar tools. Do they work better for you? Are there trade offs?

      Obviously AI can comment on style and structure, I just hadn't thought about grammar before as a use case.

      • IshKebab 21 minutes ago
        Standard modern grammar tools do use AI. The 00s era grammar checkers were useless trash.

        But yeah I feel like they are perfectly sufficient for spelling and grammar. You don't need an LLM. But I would probably still at least see what an LLM says anyway. Occasionally they have good writing suggestions and it's not really any effort to try.

    • paganel 53 minutes ago
      > AI check for grammar and spelling mistakes

      To the contrary, in this very sad AI age we're now living seeing genuinely human grammar mistakes is refreshing. Perfect-ness is the enemy of human ingenuity.

      Later edit: To quote an Austrian Communist Party member from back in the early '30s (even though the quote was made in the late '50s - 'early '60s):

      > As machines become more and more efficient and perfect, so it will become clear that imperfection is the greatness of man. [1]

      [1] https://en.wikipedia.org/wiki/Ernst_Fischer_(writer)

      • syshyk 31 minutes ago
        I honestly don't see that grammar mistakes that that much of a problem to start with.

        It only matters if what you are writing would be misunderstood, no?

        The rules of English are a complete mess anyway, not to mention region specific.

        I really don't care any more.

        I have started to argue that punctuation only matters if you strip it from sentences after the fact, any of these examples people give for why its necessary from lack of capitalisation can be removed by writing without that in mind.

        An on an AI specific note that bridges language... I have colleagues here on AI courses, in the UK, they have to go through some English and Maths tests. It's meant for young apprentices, a complete government subsidy grab if you ask me, even though they're a seasoned professional. There is a huge amount of Americanisms in there, and the fact it refuses a specific type of " marking it as wrong, the Oxford comma argument etc - so I wouldn't trust AI(LLMs) to be correct either.

        I am useless at writing long documents, so toyed with trying to get LLMs to reflow them without changing anything (I doubt it), but time and time again it just turns it into worse tosh.

        I cannot stand reading the long winded emails and "policy documents" that someone has AI spun up, and likely not even read. Working with people that don't put in the work is pretty shit.

      • alex43578 40 minutes ago
        Ahh, the wonderful value of human imperfection: deaths to medical errors, false convictions in courts, and the myriad everyday mistakes in judgement.

        If human ingenuity invents a self-driving car, will we long for the days of driving ourselves into a tree just because we miss being imperfect drivers?

        • xigoi 6 minutes ago
          You’re spitting in the face of the numerous people who have been falsely convicted by LLMs.
        • syshyk 28 minutes ago
          Sounds like a straw man to me?

          The quote above generally points to the fact that our mistakes and basically hunches is what drives change and innovation. And I do feel a bit sickly business saying that.

          AI is spinning up some interesting new things (yes, ill just call it things, enjoy) but I can only feel that what we're seeing is the wisdom of the crowed as the LLM averages our thoughts as they currently stand. This can only happen once.

          • alex43578 16 minutes ago
            1) I don’t see that in the quote. To me it reads as a nostalgia for human imperfection: an impressionist painting is inherently more valuable than a photograph of the scene, for example.

            2) LLMs have clearly surpassed the “wisdom of the crowd”. Their abilities are spiky, but I guarantee the latest frontier model is more intelligent than 80% of people across a number of domains. For math, given the ability to answer questions that have gone unanswered for years or decades, it’s clearly smarter than 99% of people.

            Calculators were not a high watermark for our ability to do math. Making mistakes in math doesn’t reliably move math forward for the population, and even for the individual only shows what they don’t know.

    • snakeboy 44 minutes ago
      > to bounce ideas off and catch any glaring errors

      These are two wildly different things. Microsoft Word's spell checker can catch glaring errors, but "bouncing ideas off" of an AI sounds more akin to passing your ideas through a "consensus sieve".

    • hnarayanan 44 minutes ago
      I found this particular typo quite funny.
    • AirMax98 52 minutes ago
      > how hard is it to right 10 pages yourself

      Tough one.

    • budman1 50 minutes ago
      that spelling mistake was done on purpose, i expect.
    • dguest 53 minutes ago
      [dead]
    • revolution88 1 hour ago
      Shame that he made a mistake? :) Better use AI and read slop? :)
      • pascahousut 53 minutes ago
        I wish a healthy middle ground was an option.
      • RobotToaster 51 minutes ago
        Grammar checkers have been around a long time before AI. (At least if you ignore that fuzzy logic used to be called "AI")
  • Xinti 38 minutes ago
    You know that this whole nonsense complaining and heel dragging against llms will absolutely fail right? The entire industry can collapse tomorrow and the tech will still be there. You can't convince people that shitty repetitive, pointless, and soul killing nonsense is somehow "character building". There is no technological advance that has ever been stopped because people whined and complained: maybe adapted, but if you are a buggy whip manufacturer: probably should pick another industry.

    Every fucking time I hear "I can tell that ai wrote this" and then it gets pilloried I get closer and closer and closer to creating a self improving hardcore red teaming exercise against that bullshit thinking. Literally there is no fucking right to know who wrote something and be free from ai. Zero. None. And the more you think you are smarter than that: the more that you deserve your comeuppance.

    • xigoi 3 minutes ago
      > There is no technological advance that has ever been stopped because people whined and complained

      Leaded gasoline?

    • qmr 35 minutes ago
      I find this comment gross.

      ... and when the AI becomes an Ouroborus?

      • Xinti 28 minutes ago
        Easy. Input from outside. "Oh gadzooks! This is written by an ai"... inject mutation into the process. genetic programming. Use throwaway accounts. And the very best part? It's as easy as a codex subscription for 20 bucks a month. Gross? Cool. Your opinion is noted.

        Way before your time, my time, or your parent's time there was a wonderful saying "starving artist" that's why people aren't artists (read people doing things they demand to do by themselves with some "handcrafted" manner.

        Yup. Sorry it's not a democratic vote about how people express themselves; nor, is anyone entitled to "out" someone else. I love how it's all down to nonsense and vibes in these subjective assertions. Anything that is codified can be worked around.

  • nickm12 15 minutes ago
    So was it a complement or an insult? I feel like I've been left hanging.
  • athrowaway3z 16 minutes ago
    I see its another entry in the long list of "use the tool for just the right amount at the right time" plea.

    Its not that i disagree with the author about his worries; its just that i've read the same personal opinion & anecdote more than a dozen times, and they all seem to disagree where the line is.

    'Our blessed productivity multiplier, their barbarous lack of thinking'

  • Xinti 18 minutes ago
    And this nonsense discredited idea:

    "Think about what the creators of the content you consume want you to think2. Treat the things you consume like the Year 7 English class where you were taught how to identify persuasive techniques. Think about things as much as possible."

    You assume that there is some Machiavellian scheme that is deliberately orchestrated like the Illuminati to do...stuff...and things...

    Businesses want to make money. Anything you see around that would be network effects of doing so. It's not a fucking scheme that someone planned out like some nonsensical conspiracy theory: companies are people too (more specifically composed of people).

    Okay, I'll bite: you presume, that you go from "read my artisanal science report" to "yeah man... like ideas come from people...from THE MAN...man...". Cool I can tell your nonsense just ran off the track of your own terrible thinking and lack of focus: nice, now I can tell how they could tell you wrote things.

    You're a scientist right? Cool beans. Prove to me. absolutely PROVE TO ME, that your approach is somehow better, and by better I mean statistically better, you know high p values, peer reviewed, most citations, cohort studies, etc.

    I'm betting that you could try to prove this (literally) personal opinion of yours and wouldn't get anywhere. I've heard the argument and no one has any level of proof approaching this.

  • snvzz 26 minutes ago
    Next sentence is likely: Should have had the AI review it before sending it to me.

    And not wrong. Because there is no excuse to present something riddled with mistakes an AI would have found in an instant.

  • atoav 30 minutes ago
    I also write documents like that myself and have been often praised for the quality of how I structure my thinking and writing. Having studied philosophy, worked as a programmer and freelanced as a graphic designer, I have an intersection of experiences that results in clearly structured documents.

    Now since AI has reached a certain quality level my worry is that what I would have written normally before comes across as AI written to the uninitiated. It started with the em-dash — a piece of typography I often used, until it came to infamy as a tell on AI-written texts.

    I still think AI often makes bad taste decisions, regardless of whether it is around text or graphic design. However I have to conceded that there are not many out there whose own decisions were better. But at least they were their own. I think as a consequence people will become even more lonely and isolated, since now you can't even be sure that the words a person chose to write to you, were their own.

  • Rodmine 4 minutes ago
    [dead]
  • knotruby671 16 minutes ago
    [flagged]
  • alex43578 44 minutes ago
    I don’t get this knee-jerk association of AI = lies or wrong. AI-written content can certainly make mistakes, but so can human content.

    The author says “That you're disadvantaged by not using it… Apparently actually knowing things doesn't matter nearly so much as knowing how to prompt a machine to do the thing.”, which is hilarious coming from someone taking an engineering degree.

    Just like how CAD and FEA have supplanted hand-drafting and manual calculations, so too will LLMs supplant hand-writing a report.

    Newsflash: you will absolutely be disadvantaged by not using tools to advance and accelerate your work. For FEA, you need to understand and validate the results - you don’t get extra credit for doing the math (or in the case of writing, doing the writing) by hand. Use an LLM, understand how to prompt it, validate/read/refine the results.

    • throwaway19268 34 minutes ago
      LLMs were built on the sum of decades and decades of manual human effort and knowledge. What you're talking about is exactly the delegation of thinking the author is talking about. I ask to what end the reliance of AI leads, it's clear LLMs lack the same level of creativity and ingenuity of real humans, and even if they were on par, if we've delegated all the thinking away then who's to say we'll even understand the results. There is a _massive_ difference between using CAD tools to remove the need for manual calculations vs prompting an AI to write an entire report on a subject you have zero understanding of. And I believe you cannot truly build an understanding without at least a moderate amount of effort.
      • alex43578 24 minutes ago
        “It's clear LLMs lack the same level of creativity and ingenuity of real humans” is completely unsupported, particularly in light of LLMs directly generating answers to questions humans have been unable to solve for years.

        I’d agree that simply prompting “Write a 10 page report on helicopters” is going to leave them without understanding. But providing the data, refining the output, and adversarially evaluating the output of a computer are all essential skills for a modern engineer. There is no value in effort for effort’s sake. We don’t make students learn everything from first principles and force engineers to develop their own equations just because it would be a “delegation of thinking” to build on the work of others, whether it be from a textbook or LLM.

      • yunwal 25 minutes ago
        > prompting an AI to write an entire report on a subject you have zero understanding of

        It continues to astound me all of the people who insist on turning off their brain just because they’re using one tool and not some other tool. Like, did you turn off your brain when you first used a computer to solve a math problem? You’re still dealing with a prediction machine. It’s a very very powerful prediction machine, but it’s a prediction machine nevertheless. So just… don’t stop thinking when you use it? You literally don’t have to interact with an LLM this way, no one is telling you to have it write your paper for you end to end, you are doing that to yourself.

  • hypfer 27 minutes ago
    What workflow requires a report with a fixed length on an arbitrary topic?

    The idea of writing documents of course makes a lot of sense, and it is also true that they probably should end up somewhere along the length of 10 pages, if the topics are all somewhat equally shaped and sliced.

    But "do a report and make it 10 pages" without context to me sounds exactly like the kind of bureaucratic assignment that of course gets offloaded to LLMs. Any sort of bureaucracy where the purpose isn't apparent, is indirect, or might not make sense, will be automated by LLMs, because no one (rightfully!) wants to do that.

    So it might be a process and management failure when people do not see why the task is actually necessary. And, if the majority of the reports handed in is written by AI, maybe it isn't?

    ___

    I guess the takeaway here should be that the heuristics previously used as performance optimizations are broken, and it is now time to update them. Which is probably not easy, but just because alternatives are hard to find doesn't mean that the status quo has not been bullshit.

    • petterroea 13 minutes ago
      These kinds of requirements are usually used in academia to force you to go enough into detail, etc. Useful when you are dealing with people who don't want to do the task or are bad at writing, although a bit arbitrary. Maybe a necessary evil for some, but with increased maturity you'd expect to not have to ask for a minimum.

      Not everyone understands that writing is one of the skills they are being taught

      • hypfer 8 minutes ago
        Yes, of course. This is clear to every reader.

        But it's trying to doctor with symptoms instead of focusing on the root cause that is lack of intrinsic motivation.

        Why are people in academia that don't really want to be there?

        I believe that any human worth keeping around and alive (read: all of them) has intrinsic motivation. Why are we not seeing it?

        I want to see that answered and fixed instead of people just cursing at the inevitable war of workarounds

        ____

        Even a stressed student juggling multiple jobs to stay afloat will care about something. And that care will drive them.

        Why are we not harnessing that care and intrinsic drive?

        Why are we instead throwing them something that vaguely resembles a vector pointing at a desirable outcome if some preconditions are true and then get angry when that obviously doesn't work for all.

    • azornathogron 16 minutes ago
      The author is at school/university. Report writing with a fixed length and somewhat arbitrary topic is a completely normal task to give to students, because the point is for them to get practice writing reports.
      • hypfer 14 minutes ago
        Yes, but it's also complete bullshit. That is exactly my point.

        Not the idea of course. The idea of repetition and forced engagement is indeed how learning happens, and you gotta facilitate that somehow.

        But that somehow has so far been mostly terrible, because most people kinda suck at didactics and sparking genuine intrinsic motivation to understand something.

        So instead you get the crushing force of bureaucracy, which, of course, gets side-stepped, the moment that becomes viable.

        ___

        Of course this is only in part on the people teaching, and a huge chunk of it is also the system forcing people into academics, who actually would be better off literally anywhere else.

        Academics came up with blunt defenses against that, and the other side now adapted too.

        The root cause however is to be found elsewhere.

  • Xinti 33 minutes ago
    God this is pretentious.

    An idiot wrote:

    "But sitting in a room telling an AI "ignore the approaching deadline, we still have 8 hours, keep working" isn't exactly filling me with enthusiasm for the rest of my career.

    It's soul crushing, and I don't want to outsource so much of my thinking to it."

    Welcome to WORK. It's not called FUN. It would be the height of ignorance to assume that you can really not do things you hate. Anyone who is 30 or older will realize that you are deluded.

    • Geventax 30 minutes ago
      But this isn't work, this is education. The important part of education isn't to end up with a product, but to learn. And if you are told to outsource the learning then what is uni even good for?
      • Xinti 26 minutes ago
        Learning was getting to the point where this person was doing some kind of aerospace/mechanical engineering to even write this report.

        Writing the report was the work. Different things.

  • TheLudd 51 minutes ago
    > how hard is it to right 10 pages yourself?

    Is this a joke?