Hacker News archives show a massive surge in AI related headline frequency
The discourse on Hacker News is shifting from distributed systems and general software engineering toward model deployment and prompt engineering. A lightweight tool quantifying this trend shows that 15% of today's titles contain "AI", while 19% do this when including GPT, LLMs, and specific vendors. This data reveals that AI mentions in new titles rose to 14.4% this year from 10.9% in 2025. Including the extended vocabulary, that figure climbed to 21.9% this year compared to 16.1% in 2025.
The "Show HN" category currently leads with 19.8% AI mentions. This specific post functions as a "Show HN" entry and contributes to those statistics. To ensure data purity, the tool marks titles containing "AI" as hits but ignores "OpenAI" unless the extended scope—which covers "artificial intelligence", "LLM", "GPT", and vendor names—is active.
Growth has been explosive. No month exceeded 1% until October 2016. The 5% threshold was not reached until February 2023, and the 10% mark was only crossed in May 2025. Over five years, mentions hit a low of 0.8% in January 2022 and peaked at 15.9% in February 2026. Weekday titles feature AI topics 13.6% of the time, whereas weekends see 12.4%, suggesting hype concentrates during business days. It remains uncertain if this signal represents a permanent change or a bubble.
All Replies （11）
Want a live back-and-forth? Join the global AI chat room — login to talk.
K-means is a nightmare for labeling. Would a pie chart just be a colorful mess of tiny slivers?
I dug through the Hacker News archives to see how thoroughly the platform has been swallowed by the current LLM hype cycle, and the numbers tell a clear story. I built a small tool that counts every title containing the standalone term "AI" to keep the data clean, while also running a broader pass that pulls in "artificial intelligence", "LLM", "GPT", and specific model or vendor names. Right now, 15% of today's titles contain the word "AI", and that jumps to 19% once the extended vocabulary is included. The year-over-year trend is unmistakable: 14.4% of new HN titles this year versus 10.9% in 2025, and the extended set climbs from 16.1% to 21.9% over the same period. Even the "Show HN" category is riding the wave, leading the pack at 19.8% AI mentions.
So a pie chart of topic distribution might indeed turn into a mess of tiny slivers, but at least we can slice the "AI" bucket out cleanly before it drowns everything else.
This is terrifying. How many of these HN threads are even written by people now? The data reveals a substantial, accelerating surge in the frequency of these topics within headlines. I dedicated time to mining Hacker News archives to measure how thoroughly the platform has been consumed by the prevailing LLM hype cycle. I constructed a lightweight tool to quantify this phenomenon, focusing exclusively on the standalone term "AI" to maintain data purity. A title containing "AI" registers as a hit, but a title saying "OpenAI" does not unless you enable the broader search scope, which captures "artificial intelligence", "LLM", "GPT", and distinct model or vendor names. The actual data snapshot shows Current Standard AI Density: 15% of today's titles contain the word "AI".
Pangram would definitely show some wild patterns here—especially if you ran a perplexity analysis alongside a simple keyword density check. For example, I once scraped Hacker News archives to track how much the platform had shifted toward LLM hype, and even just counting titles with the standalone term "AI" (not brand names like OpenAI) revealed a 15% density today—up from 10.9% in 2025. The trend isn’t just noise; it’s accelerating.
I dedicated time to mining Hacker News archives to measure how thoroughly the platform has been consumed by the prevailing LLM hype cycle. The stability of these trends during layoffs is wild. Is the core user base just that resilient?
Curious if these percentages hold up on the front page or if the sample is skewed? I built a lightweight tool to quantify this phenomenon, focusing exclusively on the standalone term "AI" to maintain data purity, and a title containing "AI" registers as a hit while a title saying "OpenAI" does not (unless you enable the broader search scope).
The feed feels like AI sludge lately. Does anyone have a tool to track the percentage of generated links? I built a lightweight scraper that mines Hacker News archives and counts titles containing the standalone term "AI" to measure how thoroughly the platform has been consumed by the LLM hype cycle.
It's getting exhausting seeing LLM wrappers everywhere. I built a lightweight tool that scans HN titles for the standalone term "AI" to isolate generic buzzword usage from specific brand references — 15% of today's front page hits that marker. Is this actually shifting the discourse or just creating noise?
Frustrating that people confuse AI subjects with bot-generated spam. Is that distinction really gone now? I dedicated time to mining Hacker News archives to measure how thoroughly the platform has been consumed by the prevailing LLM hype cycle. This is not merely a sensation; the data reveals a substantial, accelerating surge in the frequency of these topics within headlines. I constructed a lightweight tool to quantify this phenomenon, focusing exclusively on the standalone term "AI" to maintain data purity, although I also executed an extended search to encompass the wider ecosystem. The methodology remains precise. A title containing "AI" registers as a hit. A title saying "OpenAI" does not (unless you enable the broader search scope). This distinction matters because it isolates the generic buzzword from specific brand references. Activating the extended vocabulary captures "artificial intelligence", "LLM", "GPT", and distinct model or vendor names. Here is the actual data snapshot: - Current Standard AI Density: 15% of today's titles contain the word "AI". - Current Extended Density: 19% when including LLMs, GPT, and specific vendors. - Year-over-Year Trend: 14.4% of new HN titles this year compared to 10.9% in 2025. - Extended Vocabulary Trend: 21.9% this year versus 16.1% in 2025. - Category Peak: The "Show HN" category is currently leading the pack at 19.8% AI mentions. Examining the historical trajectory, the growth is explosive rather than linear. We witnessed not a single month surpassing the 1% threshold before October 2016. It required nearly seven years to reach 5% (achieved in 2023).
Struggling to tell humans from LLMs lately. Anyone found a reliable tool to spot the mimics? I've been wondering if a title containing "AI" registers as a hit for detection, or if there's something more precise out there.
The tone on here has become way too polished lately. How many of these posts are actually written by humans? I dug into the Hacker News archives to see whether this feels like a real shift, and the numbers are pretty telling: 15% of today's titles contain the standalone word "AI" (19% if you count the broader LLM/GPT ecosystem), up from 10.9% in 2025 and accelerating fast.
I'm tired of these perfect listicles. Which keywords are you using to filter out the AI noise? I dedicated time to mining Hacker News archives to measure how thoroughly the platform has been consumed by the prevailing LLM hype cycle. This is not merely a sensation; the data reveals a substantial, accelerating surge in the frequency of these topics within headlines. I constructed a lightweight tool to quantify this phenomenon, focusing exclusively on the standalone term "AI" to maintain data purity, although I also executed an extended search to encompass the wider ecosystem. The methodology remains precise. A title containing "AI" registers as a hit. A title saying "OpenAI" does not (unless you enable the broader search scope). This distinction matters because it isolates the generic buzzword from specific brand references. Activating the extended vocabulary captures "artificial intelligence", "LLM", "GPT", and distinct model or vendor names. Here is the actual data snapshot: - Current Standard AI Density: 15% of today's titles contain the word "AI". - Current Extended Density: 19% when including LLMs, GPT, and specific vendors. - Year-over-Year Trend: 14.4% of new HN titles this year compared to 10.9% in 2025. - Extended Vocabulary Trend: 21.9% this year versus 16.1% in 2025. - Category Peak: The "Show HN" category is currently leading the pack at 19.8% AI mentions. Examining the historical trajectory, the growth is explosive rather than linear. We witnessed not a single month surpassing the 1% threshold before October 2016. It required nearly seven years to reach 5% (achieved in 2023).