How much of HN is AI?
AI now dominates Hacker News to the point where users estimate roughly half of front-page stories are either about AI or AI-generated, prompting both fatigue and concern about authenticity. Commenters debate whether this reflects AI’s genuine importance or just another hype cycle, compare it to past obsessions like crypto and NoSQL, and worry about bot-written comments, astroturfing, and the loss of “hand-crafted” technical content. Some propose filters, alternative sites, or smaller curated communities as ways to escape the AI deluge and preserve more human-centric, exploratory tech culture.
Perceived effectiveness of AI detectors (Pangram)
- Some commenters consider Pangram highly effective, citing third‑party evaluations claiming ~90% detection and very low false positives.
- Others report low confidence, saying it produces both false positives and negatives.
- Critics note many “evaluations” are collaborations with Pangram and may be cherry‑picked; older work arguing “AI detectors don’t work” is mentioned but may be outdated.
How much of HN is AI‑related
- Multiple rough estimates: 40–60% of front‑page stories are about AI or clearly AI‑generated, with several users informally tracking 4–6 out of top 10 posts as AI‑related.
- Tools filtering AI from HN (browser extensions, custom readers, unslop.news, etc.) typically report roughly half of the top stories removed.
- Some suspect moderation might implicitly cap or stabilize the proportion; others just see it as reflecting current industry focus.
Fatigue, hype, and “appropriateness” of AI dominance
- One camp argues AI/LLMs are the biggest tech development in decades (or ever), comparable to early internet or smartphones, so heavy coverage is expected.
- Another camp experiences severe fatigue, comparing AI to past hype cycles (crypto, NFTs, metaverse, NoSQL) where discourse and quality suffered.
- Skeptics question real‑world impact: for many, LLMs aren’t part of daily life or don’t provide enough value to pay for; some still regard them as toys or “very crappy technology”.
- Optimists point to strong usefulness in coding, math, research, and security tooling, and see rapid capability growth; pessimists warn about plateaus, unsustainable compute, and weak business models.
Impact on HN culture and content
- Long‑time users describe a shift from eclectic, “hand‑crafted” hacker projects and entrepreneurship toward repetitive AI marketing, shallow opinion, and culture‑war‑style debates.
- Many express nostalgia for earlier eras (pre‑AI, pre‑crypto, JS‑framework days, “Erlang day”), and feel “old‑school hackers” have moved to invite‑only forums, IRC, Lobsters, etc.
- Others argue some of this is aging and changing perspectives; HN visually and structurally hasn’t changed much, but the audience and volume have.
Bots, AI‑generated comments, and moderation
- There is widespread suspicion of AI‑written comments and astroturfing, especially on AI topics; some believe a large share of comments are bots, others are unconvinced.
- Moderation uses some LLM detection and relies heavily on user flags; AI‑generated comments are formally disallowed.
- Users disagree on how easy it is to spot AI text; some claim it’s trivial “with the right eye”, others say they can’t reliably tell at all.
Diverging attitudes to using AI tools
- Enthusiasts claim almost everyone in tech now uses LLMs heavily, especially for coding, and that they can function like cheap personal assistants.
- Others say most colleagues avoid them even when licenses are provided, or use them only occasionally with mixed results.
- Some mourn a perceived loss of craftsmanship and fun in programming when “vibecoding” replaces writing code by hand; others see LLMs as just another tool that shifts what counts as skilled work.
- Environmental and local‑impact concerns (energy, water, noise, centralization) are raised; supporters often see these as overstated or solvable externalities.