March 11, 2026
Is this comment written by a toaster?
How much of HN is AI?
Is Hacker News just bots now? Old-timers shrug, creators panic
TLDR: A blogger says AI stories—and even AI-written posts—are crowding Hacker News, backed by a detector scan. The comments split: veterans say “same vibes, new topic,” while creators fear becoming bot fodder and others warn we’ll all start writing like machines—raising big questions about trust, incentives, and the future of online voice.
Is Hacker News turning into a robot-run billboard? That’s the charge after a blogger tallied February’s top posts and found AI stories dominating the feed — then used a detector called Pangram to flag pieces that looked machine-written. Cue fireworks in the comments, where half the crowd shrugged and the other half spiraled.
Old-schoolers like cj insisted nothing’s changed: topics rotate (crypto yesterday, AI today) and the signal still beats the noise. Pragmatists said just add an AI tag and scroll on by. But creators like ljhsiung went full existential, worrying their words are just free fuel for chatbots and wondering if paywalls are the last human refuge. Builders confessed they already use AI because it’s “too convenient,” while pdp dropped the creepiest prediction: we’ll all start writing like the bots.
Then came the snark. deepsquirrelnet cracked that the Turing test is dead if we need machines to tell us what machines wrote. Others joked HN now stands for Hype Neural, and someone asked, “clap if you’re human.” Meanwhile, the OP’s mention of vendor “submarine marketing” added spice — is the feed getting gamed? Whether you think it’s business as usual or the end of human vibes, one thing’s clear: the bots are in the building.
Key Points
- •The author sampled Hacker News’ daily top five stories for February 2026 to assess AI-related prevalence.
- •AI occupied four of five top slots on Feb 4 and Feb 12; Feb 5’s lineup was described as entirely AI-focused.
- •Days without AI in the top five were Feb 1 (first AI at #7, #9), Feb 9 (first AI at #8), and Feb 25 (AI at #6, #9, #10).
- •The author used the Pangram detector to identify likely AI-written stories and then manually reviewed flagged items.
- •The author reports that Pangram’s flags generally made sense and suggests there were some false negatives; a Feb 19 #3 story was given as an example of flagged content.