A tester says using AI to write your work poisons your credibility for good
Claim AI merely "helped" with your writing, James Bach says, and he will read everything you produce afterward as slop. There is no verifying how much you actually did.
His effort argument is a picture: AI-writing is hooking a motor to a stationary bike and calling it exercise, or taking a helicopter to the summit of Everest.
His trust argument is harsher. "AI helped" is unverifiable, so it's a lie about how much you did. Admit it once and all your future work reads as spam.
The thread's cleanest counter pointed out that he asks you to take his own no-AI claim purely on his word.
Two arguments, one of which eats the other
James Bach, who founded the software-testing consultancy Satisfice, argues that you should never use AI to write something you present as your own, and that you shouldn't even soften it to "AI helped." His first case is about effort. Letting a model write for you is "hooking a motor to a stationary bike and calling that exercise," or helicoptering to the top of Everest and claiming the climb.
His second case is about trust, and it does the real work. Saying AI merely "helped" is unverifiable, so it functions as a lie about how much of the labor was yours. Admit AI-assisted writing once, he says, and he will treat all your future output "as if it is slop or spam." He leaves a narrow exit for genuine, transparent co-authorship where both parties did observable intellectual work, but his conclusion is that reputation rests on demonstrated trustworthiness in an AI-skeptical world.
The thread's first jab went straight at the byline. The consulting and writing trades have lived with ghostwriters and editors forever:
"ghostwriter says not to use ghostwriter more at 11" — a reader flagging the author's own profession
The sharpest reversal turned the trust argument on its author. Unverifiability, it turns out, cuts in both directions:
"using author's logic, I should not trust them when they say they never use AI for writing, because all we have is their word." — a reader turning the argument back on itself
The highest-karma voice in the thread rejected the effort framing outright. What a reader buys is signal and individuality, not sweat:
"I don't care whether my favorite author sweated for months … or he effortlessly dictated the final form … I use the author's name as a proxy to judge whether to invest that effort myself." — the highest-karma commenter
And one reader described a disclosed, mundane use the absolutist rule simply can't hold:
"I'm not ashamed to say that I used it last week in a chat conversation with a recruiter … This recruiter didn't deserve my authentic[ity]" — a reader describing a disclosed everyday use
The split in the thread is over what a name actually certifies. Bach treats it as a promise about how the words were made. His critics treat it as a proxy for whether the result is worth their reading time. If it's the latter, then a disclosed recruiter reply and a six-month novel land in the same place, and the effort argument never had to be the load-bearing one.
Get breakdowns like this every day
The top Hacker News stories, unpacked — by email, Telegram, or RSS.
How a scary AI-hacking claim traveled general to senator to print
A model that "broke into almost all of our classified systems, not in weeks, but in hours." The line is everywhere. Trace it back, though, and nobody who said it touched a keyboard.