Live data from Hacker News

The whole premise of checking for human writing is daft

mohammedshehu.com

71–80 of 83 posts

Re: The whole premise of checking for human writing is daft

#71
post #9

I think we're just in an transition point. When I get materials that are clearly AI generated, I look at them as if there was less effort put into it. But that is irrational. If it was verified for accuracy, and edited for conciseness, and addresses the task at hand, does it matter?

Except it wasn't verified for accuracy. It never is.

The whole point of using AI to write for you is that you don't care to or can't be bothered to do any significant work yourself. If you care enough to do the research required to ensure all of the AI's output is accurate, that's like 80% or more of the work done right there, so why avoid the remaining 20%?

Nobody talks about using AI to gain a 20% productivity boost. They brag about 100% or even 1000%, which only makes sense if you're not doing any of that time-consuming "verification" at all.

Re: The whole premise of checking for human writing is daft

#72
I don't think the Stephen Hawkings example is compelling. No one argues against AI for transcription or to help paralyzed people communicate. Stephen Hawkings still wrote every word of his book.

But it does raise the question: what if he hadn't? What if Stephen Hawkings had secretly used an LLM to write A Brief History of Time? Imagine that we find a letter in which he confesses that he used an LLM to write the book. Would that lessen the value of the book? I don't think so.

I think we need to distinguish between two arguments:

A. You should avoid AI writing because it is bad. B. You should avoid AI writing because it isn't human.

I find myself in camp A. I think a lot of AI writing is bad, but (1) it gets better when a human spends time working on it, and (2) AI is improving over time and its writing is getting better.

For me, an AI writing detector is about curation. It is hard to figure out what's worth reading, so filtering out low-effort AI writing is a good heuristic.

But what I really want is a Good-Writing Detector. I don't care how a piece of writing happened--all I care about is if I'm going to value reading it.

In contrast, camp B is never going to be happy with AI writing. It doesn't matter how good it gets; even if it is better than human writing, camp B will not approve.

I get that, even if I don't agree. There is virtue in sacrificing to support an ideal.

Re: The whole premise of checking for human writing is daft

#73

Sorry but I think this essay is largely wrong. First, there is tons of evidence that working your thoughts on paper, word by word, helps develop the idea and exercises the brain. Secondly, in other endeavors we have defined clear lines between human and machine labor: chess computers are banned from human tournaments; race cars and bicycles are not allowed in footraces. And while I am very sympathetic to the advantag…

How can there be tons of evidence for a 3 year old technology? Everyone seems to think that all writers using AI are one shotting their content. A writer can carefully edit and revise their content using AI tools just as well as manual writing. That's what I do when I use AI to write. I also never use AI to write interpersonal content. Just like I won't type a Valentines letter to my wife, I'll handwrite. I only use…

Look, I understand. There is a ton of low-value writing in the world: legal boilerplate, press releases, most documentation. That writing can and will be replaced or supplemented by AI. But look up cognitive offloading. It doesn't take centuries of research to show evidence for that, and in fact the benefits of writing for thinking have been known long before AI writing became available (the alternative being just talking or thinking without writing).

Re: The whole premise of checking for human writing is daft

#74
I understand why people want you to write; because it’s good for you, because it focuses a wandering thought, because the first sentence is usually rubbish and the seventh is often better, because you learn cadence and structure and argument through repetition, because there are things you don’t know you believe until you try to explain them, and because spending an afternoon wrestling a stubborn idea into words can leave you with a stronger brain;

no, i want you to write because i want your authentic input. i don't care what you get out of it.

but none of this explains why we’ve decided that the physical production of every word, one keystroke at a time, must now serve as a test of whether a thought is human.

well, if you didn't write it yourself then it guaranteed not to be human, that's why.

It’s daft.

there is no alternative.

It takes one useful method of thinking and promotes it into the only respectable one.

it is the only one. if you didn't write it, you didn't think it. you thought the prompt, but that's not what you are sharing.

It confuses the exercise with the outcome,

the exercise IS the outcome. if you didn't exercise it, you didn't produce it.

the labour with the intelligence,

the labor is the only proof of intelligence we have.

and the movement of fingers with the movement of a mind.

if you didn't move your fingers, then your mind didn't move either.

It asks whether someone typed every word themselves when it should ask whether they had something useful to say, whether they understood it, whether they made the important decisions, whether they checked the claims,

if you didn't type (or dictate) it yourself then

you didn't say anything. you didn't understand anything. you didn't decide anything. and i have no idea whether you checked the claims.

and whether they can defend the finished work when someone else asks what on earth they were talking about.

if you had written it all yourself, maybe you wouldn't have to defend it. and if you do defend an AI text, then why didn't you write that to begin with? that was the contribution i was looking for. but then how am i supposed to know that the defense is not also AI generated?

all the same is true if you cite a lot of other peoples writing without adding anything of your own. it may be the correct message, but it doesn't tell me anything about you. but at least i know that writing came from humans and if you cite them correctly i can interrogate them directly. AI writing does none of that.

i stopped reading after the first two paragraphs. i am pretty sure i could continue refuting everything line by line.

(one more because i think i have a clever response:)

A person can type 3,000 words by hand and have no thought pass through their head beyond the need to finish before lunch.

yes, that is a possibility, even if not very likely. if you paste AI output however it is a certainty.

Another can dictate an idea, argue with a language model for an hour, reject six drafts, rearrange the whole piece, rewrite half of it, and publish something they understand down to the final comma. Under our emerging rules of literary customs inspection, the writing police see the first person as a human writer and the second as a suspicious package.

yes, you can do all that, that's called research. but it's only valuable if the final conclusion is written by yourself.

that's the same as writing a paper where you read lot's of other papers to research a topic. you can't just create an amalgamation of the texts you read. the conclusion has to be your own.

Re: The whole premise of checking for human writing is daft

#75
It’s worth pointing out that the author of this site has published ~100 articles over the span of two months. How much original thought has been put into these ideas? People are frustrated with AI slop because it’s low effort and wastes their time. They don’t care if it was written by hand, typewriter, whatever. The fact that no human effort is being expended is the problem. Giving your shower thoughts to AI is not writing.

Re: The whole premise of checking for human writing is daft

#76
post #14

Earlier quoted context omitted.

This is my thought, exactly. Reading has always been a contract between the person who wrote the text and the person who’s reading it: “these are my thoughts, recorded to share with you.” When reading LLM-generated text that purports to be a real person’s that contract is broken. Verification that even just the basic ideas are the author’s is impossible. It feels cheap and deceitful.

> Reading has always been a contract between the person who wrote the text and the person who’s reading it I don't think an author is bound to their readers via contract, implicit or otherwise. I have no obligation to tell people what tools I use, or how I've arrived at a final product. That might be upsetting to hear, but authors don't owe their readers anything, just as readers don't owe authors anything. There is…

Ghost writing is fraud.

Re: The whole premise of checking for human writing is daft

#77

Earlier quoted context omitted.

> Reading has always been a contract between the person who wrote the text and the person who’s reading it I don't think an author is bound to their readers via contract, implicit or otherwise. I have no obligation to tell people what tools I use, or how I've arrived at a final product. That might be upsetting to hear, but authors don't owe their readers anything, just as readers don't owe authors anything. There is…

Ghost writing is fraud.

I don't think a single ghostwriter has ever been convicted of fraud for writing an auto biography on behalf of another person. Do you have some evidence to back this claim or am I expected to take your word for it?

Re: The whole premise of checking for human writing is daft

#78

Why are comments turned off on the blog posts itself? I was planning to make 200 thoughtful comments with the cheapest AI I could find to prove a point, but the lack of an option to comment shows you already know the point: something has to gate participation in the conversation. Human authorship is a fantastic gate, which keeps me from pulling my stunt here, where that gate is firmly in place. You could achieve almo…

Effort is not a good signal for value. Never has been. Just because someone took 10 more hours to finish a task doesn't mean they did a better job or they care more about the result.

Re: The whole premise of checking for human writing is daft

#79

Why are comments turned off on the blog posts itself? I was planning to make 200 thoughtful comments with the cheapest AI I could find to prove a point, but the lack of an option to comment shows you already know the point: something has to gate participation in the conversation. Human authorship is a fantastic gate, which keeps me from pulling my stunt here, where that gate is firmly in place. You could achieve almo…

Effort is not a good signal for value. Never has been. Just because someone took 10 more hours to finish a task doesn't mean they did a better job or they care more about the result.

Effort is, however, an okay proxy for giving a shit.

You’re right that some folks can share more value in fifteen seconds than others might in fifteen hours (setting aside that the first set of folks might have got to that point by way of fifteen years of practice first). That’s hard to gate on. But keeping out shitposters and spammers by requiring a modicum of effort is at least somewhat effective.

Re: The whole premise of checking for human writing is daft

#80

Why are comments turned off on the blog posts itself? I was planning to make 200 thoughtful comments with the cheapest AI I could find to prove a point, but the lack of an option to comment shows you already know the point: something has to gate participation in the conversation. Human authorship is a fantastic gate, which keeps me from pulling my stunt here, where that gate is firmly in place. You could achieve almo…

Effort is not a good signal for value. Never has been. Just because someone took 10 more hours to finish a task doesn't mean they did a better job or they care more about the result.

But a lack of effort is a good signal for a lack of value.

Eleven times out of ten, your initial idea is garbage. If you write it up, or let AI do that for you, the result is garbage. But if you go through the effort of structuring the idea and thinking about how to express it and how to phrase the details, you may learn what's wrong with the idea and how to improve it.

Post reply on HN