Noah Smith shares a bonus podcast episode in which he talks with Azeem Azhar of Exponential View about the AI boom and the possibility of an AI bust. Smith says the discussion was sparked by one of his posts from a few weeks earlier, that they reached no firm conclusions, and that more discussions with Azhar are expected in the future.
A researcher analyses Jørn Lier Horst's AI-assisted novel "Den siste saken" (The Last Case), finding that sections flagged as AI-generated contain more nouns and adjectives and fewer pronouns and verbs than his pre-ChatGPT novels, along with AI-typical clichés like humming, quiet, heavy air and sliding gazes. The analysis contrasts the novel's static setting descriptions with the action-embedded descriptions of earlier books such as "Katharina-koden" (2017) and "Nattmannen" (2009), and questions Horst's claim that using ChatGPT for research and editing is no different from consulting books or locals.
An Atlantic article, shared on Hacker News, argues that preparing for an AI apocalypse should resemble preparing for a snowstorm. The submission links to the article and its Hacker News discussion thread.
An Ask HN post poses the question of how one can care about a project that is predominantly generated by AI. The submission links to its Hacker News comments page and had 1 point and 2 comments at the time.
According to Village's LLM Atlas, which scores 857 models on a single capability scale, China's best open-weight models have led America's every day since September 2024, when Meta's Llama 3.1 405B last matched Alibaba's Qwen2.5. The 12 record-setting US open models since 2025 arrived a median of 3.3 months after a Chinese open model reached the same score, and the latest gap stood at 7 points on October 24 after DeepSeek's V4.1 Flash and Moonshot's Kimi K3.1 Pro widened a lead Meta's Muse Spark Open had cut to 2.9 points. Both countries' open models still trail the best closed models, by about five months for China and nine for America.
Applied mathematicians Steven Strogatz and Alex Townsend have written a new book, Big Math: The Hidden Codes That Run Our World, examining how "big math" enabled by computers, large language models, and eventually quantum computing is beginning to outstrip its human creators. The article notes Strogatz's background, including his 1998 paper with Duncan Watts on small-world networks, his study of the Millennium Bridge wobble, and his popular books such as Sync and Infinite Powers.
An Ask HN poster says non-prepper friends are stockpiling goods citing "AI and cybersecurity," and asks whether this is unsubstantiated panic or something worth more attention. They note being deeply familiar with AI developments and cite reports of countries issuing notices and the Pentagon's 90-day off-grid plans, while saying the prepping feels "a little Y2K" to them.
Sayash Kapoor, Rishi Bommasani and Arvind Narayanan announce STABL, a new non-profit research organization arguing that AI's societal impact will be limited more by the pace of institutional and organizational change than by capability progress. STABL outlines four research directions: empirically testing recursive self-improvement claims (including open-ended evaluations via CRUX), expanding AI safety beyond alignment and non-proliferation toward control, downstream defenses and societal resilience, measuring AI's economic impacts and steering toward augmentation, and studying how knowledge work and institutions should adapt.
A YouTube video titled "My Brutally Honest Thoughts About the AI Bubble" was submitted to Hacker News, where it received 2 points and no comments. The supplied material contains only the video link and the Hacker News discussion link, with no description of the video's content.
New research from OpenAI and Upwork finds shifting demand for freelance work, new skill combinations, and larger budgets and more hours for AI-related projects.
A programmer who describes himself as anti-AI argues that denying LLMs' usefulness in software engineering is denial, and that weak arguments about AI writing bad code undermine legitimate concerns. He notes that many open-source projects ban LLM use outright, but says using AI to review code and dependencies is enough to catch vulnerabilities, since hand-written code in projects like Node.js, PHP and musl also had flaws found by AI. He urges shifting the debate to AI's effects on mental health, curiosity, creativity and the dehumanization of the craft, including communities that once rewarded tinkering and hack value.
The author collected about 800 Hacker News comments from 2016 to 2026 of the form "wake me up when AI can…" and "I'll believe it when…", then let readers vote on whether each bar had been cleared; over 100,000 votes from almost 9,700 people were cast before voting closed. Of 826 statements, 112 were judged "clearly happened" (mostly code and text tasks), 89 "clearly not yet" (mostly physical-world and novel-discovery tasks), and 78 were contested.
ActivTrak's Productivity Lab reports that AI users have workdays up to 90 minutes longer than non-users. The finding is published at activtrak.com and discussed on Hacker News.
The author argues that LLM coding has made it easy to generate projects that appear well-tested, documented and organized, so traditional selection criteria like test results, code coverage and documentation no longer signal quality. With the number of such projects growing, evaluating options takes far more effort, and community size is often the only remaining criterion. The author lists tools such as Slopscore, Antislop and Stopslop for spotting LLM-generated code and calls for repository badges or an "AIness" score.
A first-person WIRED essay describes the "sad wives of AI": partners of men consumed by AI work or ambition, who say the industry's hours and obsession strain their relationships and leave them with unpaid emotional labor. The author cites Rutgers labor studies chair Yana van der Meulen Rodgers on the gendered "ideal worker" trap and therapists who say some wives now process marital conflict with chatbots. She notes she did not interview any of the husbands for the story.
The article argues that Apollo and Ares research undermines the AI revenue narrative and the legal protections behind AI-related debt, citing Apollo Chief Economist Torsten Slok's finding of little evidence of corporate productivity gains and Ares' review of actual deals showing off-balance-sheet structures, weak guarantees with contractual "outs," and data center property insurance covering only a fraction of asset value. It notes Apollo's roughly $80B in AI debt commitments for 2026 despite these warnings, and cites ING seeking to offload data center loans, Microsoft terminating an Oracle AI cloud commitment, and an unverified Musk claim that SoftBank defaulted on a gas turbine contract.
An Ask HN post asks for opinions on companies that are actually making money by using AI, either by granting access to Claude or by including some kind of AI feature in their products, noting that AI news tends to focus on the same companies improving their models. The post links to its Hacker News comments page.
Jagular.io's AI Heat Map publishes a daily 0–100 reading of how stretched the AI boom is, built from 19 indicators drawn from free public sources including FRED, SEC EDGAR, Yahoo Finance, the npm registry, Y Combinator's company list and the Ramp AI Index, with 50 as normal and the dot-com peak as the yardstick for 100. Pillars are weighted Prices 30%, Borrowing & risk appetite 20%, Spending strain 20% and Speculation 15%, with real demand taking the remaining 15% as a counterweight. A separate noise meter tracks AI-bubble talk over the last 28 days via Hacker News mentions, Wikipedia views and prediction-market volume, and the author notes the project is not investment advice.
Azeem says the past two weeks have been chaotic in tech, citing a correction in AI companies, the market shock of the Democrats' mid-term victory that gave the Nasdaq 100 its worst day since March 2020, and news that Jensen Huang has a relationship with a Chinese spy. He argues this is a flight to quality rather than a collapse: spending keeps rising, AI revenue reached $188bn in the twelve months to October and runs at $345bn a year, and Anthropic's growth falling to 45% a quarter puts it in line with the market. He expects Democrats to make data centers dearer, favoring hyperscalers, Nvidia and the open-weight camp to have their wings clipped, an Anthropic IPO at a sensible price to steady the sector, and neoclouds to be the weak link.
Paramount Skydance CEO David Ellison said on an earnings call that AI could have a bigger impact on film than CGI, comparing it to the early-1990s advent of CGI and digital composites. He said the company has been developing its own AI tools that could make budgets more digestible and enable more original storytelling, while stressing that AI is a tool for artists and "will never be a replacement for them."
A Hacker News submission links to a post by @zhizdev on X asking why AI-generated "slop" is so frustrating to read, with the discussion thread on Hacker News. The supplied material contains only the link, points and comment count, not the post's content.
A Gizmodo article argues that The New York Times misunderstands how AI contributes to suppressing wages. The item links to the Gizmodo article and to a Hacker News discussion thread with 4 points and 4 comments.
The post argues that AI labs keep per-token prices unchanged while increasing output tokens per task, citing GPT Image 2.6 using 3.3x the output tokens of 2.5 on the same image and Anthropic's tokenizer change producing about 30% more tokens for the same text. It adds a November 9 update that Claude Fable 5.2 is cheaper per token than 5.1 ($8/$40 vs $10/$50 per million) but cost 3.4x as much on the same drawing, and that Opus 5.5 used 1.8x the output tokens of Opus 5 at xhigh. The author says Playcode keeps 2.5 as default and suggests comparing usage.output_tokens on the same prompt.
An analysis of 12,000 developers across 200 companies in Q1 2026 found that the median AI coding user consumes about 51 million tokens per month (roughly $52.38), while 90th-percentile users consume about 380 million tokens (roughly $691.14). Higher token usage correlates with more merged pull requests, but cost per PR rises from $0.28 in the lowest usage tier to $89.32 in the highest, and the median developer uses about 7 million tokens per PR versus roughly 69 million for the top decile. The author argues that broad, moderate adoption delivers more value than pushing a small group to extreme usage, and that an "agentic barrier" limits gains from simply increasing token consumption.
This opinion piece argues that AI is a distraction from capitalism's harms, pointing to the energy and water consumed by data centers, the destruction of ecosystems, and the use of AI like Palantir's Maven in military kill chains. It claims most AI companies lose money while NVIDIA profits, that AI erodes independent reasoning and labor, and that communities are organizing against data center expansion. It cites the game studio Hinterland's statement that there is no ethical application of AI because commercial models are built on industrial-scale theft and environmental destruction.
Reporting from Data Center Alley in Ashburn, Virginia, the author argues that public opposition to data centers is driven more by electricity than by the culture war framing Big Tech favors, since facilities drawing 54 MW or more strain local grids and force new transmission costs onto utilities. Because AI-scale sites ideally need around 6,000 MW, companies are increasingly building behind-the-meter natural-gas turbines and diesel backup in rural areas, which still affects surrounding communities through air pollution and a lack of local say. The piece notes that industry insiders see their ambitions constrained less by public opposition than by a lack of electricity, and that future sites may go where few people can resist them.
A Berkeley News article discusses how recent advances in AI are creating a 'paradigm shift' for robotics. The supplied material contains only the article link and a Hacker News comments link with 2 points and 0 comments, so no further details are available.
An Ask HN post asks whether AI is used for on-the-field sports strategy, expecting a football team to soon use it to call plays based on opponents, play-by-play records, injury and schedule details, and game situation to predict the best tactic.
A CBC article reports that businesses in British Columbia are responding to "AI slop" by using pen-and-paper advertisements. The story is linked from a Hacker News submission with 5 points and no comments.
A Hacker News user describes trying a voice conversation with ChatGPT during a morning walk to work through details of a new family project, and says it surfaced angles they hadn't considered and left them more motivated. They ask how many others on HN have voice conversations with AI, whether they find them immersive, and whether they stick to technical topics or explore other things.
An Ask HN poster says that with tools like Codex they no longer have many software problems to solve, and wonders whether this is making them intellectually lazy about finding problems. They say friends also struggled to name things they wish existed, and that most remaining work problems are process-related and interpersonal rather than technical.
The author reflects on Princess Mononoke, comparing Lady Eboshi's Irontown—which gives outcasts work and dignity while destroying the surrounding forest—to today's AI, which is genuinely empowering but carries costs borne by those who don't benefit. They note that Ashitaka, who seeks to "see with eyes unclouded by hate" and refuses to take sides, models the nuance they want in AI debates. The piece closes by asking not whether there will be iron, but what kind of Irontown we build with it.
The author distinguishes between predictions that AI will improve rapidly, which he says have been correct, and predictions that AI will transform the world, noting the absence of growth, job loss, and radical day-to-day impacts. He asks how much longer we must wait for these disruptive changes.
The author argues that AI does not create a new kind of fraud but energizes existing scams by making them faster, cheaper and easier to run at scale, drawing on three presentations about AI and fraud. He describes the anatomy of a scam (contact point, victim, continued interaction, confidence), explains tactics like the "spaghetti test," repetition and information gathering, and notes that AI voice systems can screen calls and that voice cloning enables vishing. He recounts a door-to-door impostor who used details from an old alarm-company data breach and possibly AI-generated talking points, and advises never confirming details, cutting off contact and verifying independently.
A CEPR VoxEU column argues, using 140 years of data on new jobs, that fear of AI-driven job displacement is overstated and identifies what should be worried about instead. The item links to the article and a Hacker News discussion thread.
With AI demand outpacing supply and domestic data center construction slowing, the US is projected to become a net importer of tokens between 2028 and 2030, depending on how many planned facilities are cancelled. Current systems plus new data centers can handle roughly 2.5x annual token growth, and imports begin by 2028 if 45% of planned projects fail to connect, by 2029 at 20% cancellations, or by 2030 even if every proposed facility is built on time. The piece notes that demand has been price-inelastic so far, with short-term GPU rental prices rising from $5 to more than $13 per hour over two quarters while buyers kept paying.
The author argues that AI cheaply fakes the signals we rely on as decision shortcuts—GitHub stars and READMEs, grades and autograders, writing quality in peer review—undermining them all at once rather than through slow erosion. Drawing on signaling theory, he warns this risks a "market for lemons" in open source, education, and research, and surveys responses from trust networks and new signals to AI-assisted evaluation, guardians of trust, and rethinking the underlying systems.
AI and tech shares recovered slightly after a sell-off wiped trillions of dollars from stock values, with the Nasdaq Composite closing up 1.1% on Friday, partly reversing its 5.5% fall, while the S&P 500 rose 0.3% and the Dow ended down 0.4%. Asian and European markets fared worse, with South Korea's Kospi down 6% and the UK's FTSE 100 down 1.01%, amid concerns that the Democrat midterm victory could mean slower AI development and questions over whether heavy AI spending will pay off. US employers added 165,000 jobs in October, more than double analyst expectations, and the unemployment rate eased to 4.1% from 4.2% in September.
This Import AI issue covers Physical Superintelligence's research system, running a non-public Anthropic model, finding a Lean-verified proof of the Aaronson-Ambainis quantum conjecture; Meta Reality Labs' Tactus textile-integrated tactile and pose sensing glove for manipulation data collection; a 2016 NASA paper on bootstrapping a self-sustaining lunar industry with 12-41 metric tons of launched mass; and Core Automation's distributed training of a roughly 32B-parameter mixture-of-experts model across clusters 9,000 miles apart.
A Show HN post links to an article at elviovicosa.com/war/ that presents four scenarios to 2029, framed around whether the AI boom continues or busts and whether the US goes on a war footing. The Hacker News submission had 5 points and 1 comment at the time of the data.