0:00 / 1:29ai Open GLM model beats GPT-5.5 on hallucination rate, but HN isn’t buying the scaling claim
The post argues that Z.ai’s MIT-licensed GLM-5.2, with 753 billion parameters and roughly 40 billion active, comes close to proprietary frontier models on Artificial Analysis’s intelligence index while hallucinating far less on AA-Omniscience: 28% versus 86% for GPT-5.5, according to the article. It uses that gap, plus a coding example where GLM-5.2 recognized an impossible prompt, to argue that raw model size and test-time reasoning budgets are no longer reliable proxies for truthfulness. The discussion matters because it gets at a central AI product question: whether users should prefer maximum capability, better uncertainty calibration, or cheaper, more reliable behavior.
Discussion: Mixed — Commenters were interested in the reported hallucination gap, but largely skeptical of the article’s broader conclusion that bigger models have plateaued or get worse. The dominant mood was methodological caution: people noted that hallucination rate, overall accuracy, abstention behavior, benchmark design, post-training, and real-world task distribution are different things. (Skepticism about claims that model scale has plateaued, Hallucination rate versus overall accuracy, Training and reinforcement learning incentives against saying “I don’t know”)
0:00 / 0:22ai Switzerland pitches Apertus as a fully open, sovereign AI model
Apertus is a Swiss AI Initiative foundation model effort from EPFL, ETH Zurich, and CSCS, presented as fully open: weights, data, code, training methods, and alignment principles are documented and reproducible. The project says it is designed for EU AI Act compliance, including opt-outs, PII removal, and memorization prevention, with 8B and 70B parameter models trained across more than 1,000 languages. The significance is less just another model release and more a European push for AI sovereignty: national and regional capacity to build, audit, and operate foundation models without depending on U.S. or Chinese vendors.
Discussion: Mixed — HN is broadly supportive of genuinely open models and European/Swiss AI capability-building, but skeptical about Apertus’s current competitiveness and usability. Commenters debated whether “sovereign AI” is meaningful beyond politics, questioned compliance claims around copyright, opt-outs, PII, and deletion requests, and compared Apertus with other more mature open efforts like OLMo and Nemotron. (full open weights, data, code, and training recipes are valued, skepticism about model quality versus leading open models, sovereign AI as capability-building and geopolitical hedge)
0:00 / 0:38ai Open models look good enough to quit Claude
Andrew Marble argues that the penalty for moving from Claude and GPT-style proprietary models to open-weight models has shrunk dramatically, comparing the moment to Linux becoming viable after years of compatibility tradeoffs. The immediate trigger is Anthropic’s Claude identity-verification rollout, but the broader point is that open models are now close enough, usable enough, and more controllable for many professional workflows. The remaining hard problems are privacy, hosting trust, performance gaps, and the expense or complexity of running models yourself.
Discussion: Mixed — HN was broadly sympathetic to trying open-weight models, especially as a hedge against ID verification, safety-policy changes, outages, and opaque model behavior from proprietary providers. But commenters were split on whether open models truly match recent Claude or GPT in real workflows, and privacy concerns shifted quickly from Anthropic to the practical limits of third-party hosting, EU routing, and self-hosting costs. (Open-weight models are now seen as close enough for many coding and general productivity tasks, Skepticism that benchmarks reflect real-world performance, especially against top Claude models, Privacy, data residency, and trust in inference providers dominated the discussion)
0:00 / 1:13biotech A fifteen-minute home test for Lyme-carrying ticks
LymeAlert, a $40 at-home tick test due to go on sale in August, is designed to tell users in about 15 minutes whether a removed tick carries Lyme disease bacteria. The test grinds up as many as five ticks and uses treated paper that changes color if the bacteria are present. Its promise is faster decisions about preventive antibiotics within the recommended 72-hour window, but experts quoted in the article warn that false positives could cause panic and that the test does not detect other tick-borne hazards.
Discussion: Mixed — HN’s discussion is less about the product’s mechanics and more about the fear and ambiguity around tick-borne illness. Many commenters shared long, unresolved health stories involving Lyme, long COVID, sleep, inflammation, hormones, or mold, while others urged caution about internet diagnoses and unproven treatments. The mood is hopeful that faster tick testing could help, but wary of false positives, missed co-infections, and the broader difficulty of diagnosing chronic symptoms. (Personal accounts of suspected or diagnosed Lyme disease, Concern over chronic fatigue, brain fog, anxiety, and sleep problems, Skepticism toward fringe or expensive treatments)
0:00 / 0:50science Your brain was not built for the doomscroll
A ScienceDaily piece argues that rising news avoidance is not apathy but a predictable psychological response to a threat-sensitive brain being asked to process global crises in real time. It cites Reuters Institute data showing 69 percent of Canadians and 40 percent globally at least sometimes avoid news, along with research linking negative headlines to higher click-through rates and severe “problematic news consumption” to worse well-being. The proposed fix is not total avoidance, but bounded, higher-quality news consumption paired with concrete action where possible.
Discussion: Mixed — HN broadly accepts the idea that nonstop negative news exploits human threat-detection, but the discussion quickly turns into a debate about agency: when to unplug, when to stay informed, and whether news-driven anxiety reflects media design or unrealistic expectations about risk and politics. (Negativity bias and attention capture, Doomscrolling and social media incentives, Limits of individual agency over global crises)
0:00 / 0:23science Slow Breathing May Nudge the Brain Toward Taking More Risks
A Neuron paper reports that slow breathing, particularly prolonged exhalation, can modulate brain function and shift behavior around reward and risk. The finding matters because breathing exercises are widely used for anxiety and performance, but the study suggests their effects may include changing reward responsiveness and risk tolerance, not simply making people calmer.
Discussion: Mixed — HN was broadly intrigued and personally engaged, with many commenters sharing breathing techniques for public speaking, anxiety, athletics, and stress. The main surprise was that parasympathetic activation and prolonged exhalation were linked not just to calm, but to increased reward responsiveness or risk tolerance, which some saw as useful and others found counterintuitive or potentially over-sold. (Slow breathing as a practical anxiety and public-speaking tool, Debate over whether increased risk-taking is beneficial or concerning, Connections to vagal tone, parasympathetic activation, and reward processing)
0:00 / 0:37science A math essay argues: treat logarithms like units
The linked essay proposes thinking of a logarithm without a chosen base as an abstract object, with bases like 2, e, and 10 acting as units—bits, nats, and digits—rather than as fundamentally different functions. It extends that framing into analogies with vectors, coordinate systems, and other mathematical structures. The appeal is pedagogical: it reframes change-of-base as unit conversion, though commenters debated whether the language clarifies the concept or obscures it with nonstandard notation.
Discussion: Mixed — HN readers were engaged and often enthusiastic about the essay’s pedagogical angle, especially the idea of treating logarithm bases as units like bits, nats, or digits. But several mathematically fluent commenters pushed back on the terminology and scope, arguing that 'baseless logarithm' is nonstandard, that the essay needs clearer type-like distinctions, and that some of the later analogies may be more notational play than new mathematics. (Logarithm bases as units of measurement, Connections to torsors, Lie theory, and coordinate-free thinking, Pedagogy versus nonstandard terminology)
0:00 / 1:22software Deno is taking web apps to the desktop
Deno has previewed Deno Desktop, a forthcoming Deno 2.9 feature that packages Deno projects—from a single TypeScript file to frameworks like Next.js, Astro, Remix, Nuxt, SvelteKit, and Vite SSR—into self-contained desktop apps. It defaults to the operating system WebView for smaller binaries, can opt into bundled Chromium via CEF for consistent rendering, and promises in-process bindings, cross-compilation, hot reload, DevTools, and binary-diff auto-updates. The feature is canary-only for now, and its CLI, config, and APIs may still change before a stable release.
Discussion: Mixed — HN was broadly interested and often positive about Deno Desktop as a serious new Electron/Tauri competitor, especially for small default builds, TypeScript/Deno integration, framework auto-detection, and cross-platform packaging. The discussion quickly turned skeptical around familiar web-desktop tradeoffs: native look and feel, browser/runtime versioning, bundle size, security permissions, and whether a shared CEF runtime will really work in practice. (Interest in an Electron and Tauri alternative, Default system WebView versus optional bundled CEF, Bundle size and shared Chromium/CEF runtime questions)
0:00 / 0:19software Mining your camera roll for accidental wigglegrams
The post shows a script that scans a photo library for runs of similar images using perceptual hashing, then stitches likely near-duplicates into looping wigglegrams. The author discovered years of accidental wigglegrams in an iCloud photo library because they often take many slightly different shots and never delete them. It matters as a neat example of turning ordinary photo-hoarding into a generative archive, and the script is available on GitHub for iCloud libraries on Mac or arbitrary image directories.
Discussion: Positive — HN largely found this charming and technically fun, with lots of people sharing related tools, cameras, and personal workflows. The main caveat was that the accidental results often read more like jittery mini-movies than convincing stereo depth, and a few commenters raised accessibility and file-format annoyances around looping animations. (delight in playful photo hacking, links to related wigglegram and stereogram tools, discussion of stereo depth versus simple motion)
0:00 / 0:18software Linux finally removes strncpy from the kernel
Linux 7.2 has eliminated the kernel’s strncpy API after roughly six years and about 362 commits spent removing remaining users. The function had long been deprecated because its semantics around NUL termination are counter-intuitive, it can cause bugs, and it can waste work by zero-filling destinations. Kernel code is now expected to use more specific alternatives such as strscpy(), strscpy_pad(), strtomem_pad(), memcpy_and_pad(), or memcpy(), depending on whether the destination is a C string, a padded field, or a known-length memory copy.
Discussion: Mixed — Commenters broadly welcomed the cleanup and respected the long, unglamorous engineering effort, but the thread quickly turned into a critique of C string handling, NUL-terminated strings, and standard-library stagnation. The mood was appreciative of the kernel work while frustrated that decades-old API design still creates this much risk and toil. (strncpy as a recurring source of C bugs, praise for long-term maintenance work in critical infrastructure, debate over NUL-terminated versus length-prefixed strings)
0:00 / 0:20software TownSquare makes websites feel inhabited — and immediately tests moderation
TownSquare is a small embeddable presence layer for websites: add a snippet, and visitors appear together on the page, move around, send short messages, and interact without accounts. The concept resonated because it brings back a playful, social feeling to otherwise static websites, but the HN-driven demo was quickly overrun by spam and offensive messages. That made the launch less a pure product showcase and more a live demonstration of why even tiny social features need moderation from day one.
Discussion: Mixed — HN liked the playful, lightweight idea of adding real-time presence to ordinary websites, but the live demo quickly became a case study in anonymous abuse, spam, and moderation challenges. Many commenters said they would only use it with site-owner controls, filtering, bans, or some form of client-side or AI-assisted moderation. (Playful ambient presence on the web, Anonymous public chat quickly attracting spam and offensive behavior, Need for moderation, filtering, bans, and site-owner control)
0:00 / 0:27software Linux async I/O: epoll’s old model versus io_uring’s new one
The article compares Linux’s older epoll readiness API with io_uring’s newer completion-based asynchronous I/O model, using the author’s student-built TinyGate reverse proxy as the motivating example. The core claim is that epoll still requires separate read and write syscalls after readiness notifications, while io_uring can batch submissions and completions through shared ring buffers, reducing kernel crossings and enabling features like registered buffers, zero-copy sends, and SQPOLL. For high-connection servers, that architectural shift can matter, but the HN discussion emphasized that real-world wins depend heavily on workload, kernel support, security policy, and lower-level tuning.
Discussion: Mixed — HN liked the practical, educational walkthrough, but the discussion quickly shifted from simple epoll-versus-io_uring framing into caveats: CPU affinity, allocator choices, kernel limits, security history, and whether io_uring actually helps under a given abstraction layer. Commenters were broadly interested in io_uring’s performance potential, while warning that benchmarks, deployment constraints, and security defaults matter more than API hype. (io_uring can reduce syscall overhead through batching and completion-based I/O, Real proxy performance also depends on CPU pinning, NIC queue alignment, false sharing, allocators, and kernel networking features, Several commenters cautioned that io_uring may increase CPU use or underperform when hidden behind poll-oriented async frameworks)
0:00 / 0:18software JSON-LD for personal sites meets a skeptical post-Google web
The article explains how personal websites can add JSON-LD structured data using Schema.org, with examples for entities like WebSite, WebPage, Person, breadcrumbs, and blog posts. The pitch is that this can help crawlers understand site semantics, improve rich results or link previews, and potentially help search visibility. On Hacker News, that practical how-to collided with a broader anxiety: many readers doubt that helping crawlers still helps authors when search engines increasingly answer users directly.
Discussion: Mixed — Commenters found the explainer useful, but the thread was dominated by skepticism about whether structured metadata still benefits site owners in an era of zero-click search and AI summaries. Practical voices recommended sticking to Google and Bing’s supported structured-data documentation, using OpenGraph for previews, and treating JSON-LD as targeted metadata rather than a blanket semantic-web solution. (Structured data can help search engines understand pages, but only within narrow supported use cases, Frustration that Google and LLM search may extract value without sending traffic back, Debate over JSON-LD versus semantic HTML, Microdata, RDFa, OpenGraph, and older Semantic Web ideas)
0:00 / 0:38software Proxmox gets a fast microVM hack
A longtime Proxmox user built pve-microvm, a Debian package that integrates QEMU’s microvm machine type into Proxmox VE as a managed guest type. It skips BIOS, GRUB, and legacy device emulation, using direct kernel boot and minimal virtio devices to get Linux guests booting in roughly 300 milliseconds, with a custom kernel, tiny initrd, OCI image import tooling, and Proxmox web UI changes. The appeal is a middle ground between LXC containers and full VMs: faster and slimmer than traditional VMs, but with a KVM isolation boundary that containers do not provide.
Discussion: Mixed — HN was impressed by the engineering and the promise of container-like startup with VM isolation, especially for homelabs, coding agents, and lightweight worker VMs. But commenters were cautious about the maintenance risk of patching Proxmox internals, questioned whether boot-time gains matter for long-running VMs, and repeatedly compared the approach with Firecracker, libkrun, Kata Containers, KubeVirt, and microvm.nix. (Strong interest in first-class Proxmox microVM support, Concern that patching pve-daemon/qemu-server internals could break on updates, Debate over whether sub-second boot matters versus steady-state VM performance)
0:00 / 1:39security Loupe shows what iPhone apps can quietly fingerprint
Mysk released Loupe, a free open-source iOS and iPadOS app that displays real values exposed by public Apple APIs and explains how they can contribute to device fingerprinting. It groups signals into passive access, permission-gated data, and advanced side channels such as URL-scheme probing and Keychain persistence across reinstalls. The project is meant as a hands-on privacy awareness tool: Loupe says the values stay on-device unless the user explicitly exports them.
Discussion: Negative — HN reaction is privacy-alarmed and critical of Apple’s defaults. The discussion largely pivots from Loupe itself to why iOS does not offer per-app internet permission or stronger network visibility, with comparisons to GrapheneOS, LineageOS, macOS sandboxing, and iPhones sold in China. (Demand for per-app internet access controls on iOS, Concern that public APIs enable device fingerprinting without prompts, Skepticism that permission prompts alone solve tracking)
0:00 / 1:25startups A Startup Job, Reconsidered as a VC Fee Machine
David Newgas revisits his early job at GenieDB, a UK startup later controlled by Stuart Frost’s Frost VP, after learning that Frost was sued by the SEC following an investor arbitration over allegedly excessive incubator fees. Digging through the record, he finds testimony and internal communications suggesting GenieDB paid excessive fees and may have helped siphon investor money, forcing him to ask whether the job that brought him to the U.S. existed because of fraud. The piece matters because it connects white-collar financial misconduct to the ordinary engineers whose careers, families, and lives can be quietly redirected by it.
Discussion: Mixed — HN readers were sympathetic to the author’s personal reckoning, but the discussion was broadly cynical about incubators, outsourcing, grants, and corporate accounting incentives. Many commenters shared analogous stories where jobs, vendors, or projects appeared to exist for budget games, fee extraction, tax treatment, or legal positioning rather than durable product value. A minority pushed back toward a more forgiving view: even failed or compromised companies can still produce learning, useful work, and real life outcomes for employees. (VC and incubator fee extraction, Perverse incentives in grants and government funding, Outsourcing and contractor budget games)
0:00 / 0:28startups StartupWiki pitches a free Crunchbase alternative, but HN wants proof and better data
A maker launched StartupWiki, an early free startup database meant to feel more like Wikipedia than Crunchbase: no accounts required, simple company pages, search and filtering, categorization, and a public API in progress. The Hacker News discussion quickly focused less on the concept and more on execution: commenters reported missing startups, incorrect funding and valuation data, confusing “verified” badges, and no visible content license. The founder responded throughout, saying YC data had been scraped, AI-agent profile generation was being improved, source-link bugs were being fixed, and an “agent ledger” had been added.
Discussion: Mixed — Commenters liked the goal of a free, low-friction startup database, but the dominant concern was trust: many found missing companies, outdated or wrong facts, unclear verification, and no visible content license. The founder was active in the thread, acknowledged bugs and pipeline issues, and shipped or promised fixes, which softened the response but did not resolve skepticism about AI-generated data quality. (Strong demand for an open, free startup database, Coverage is currently thin, including missing YC and prominent startups, Verification badges need source links, provenance, and timestamps)
0:00 / 1:19policy Claude ID checks spark privacy backlash
Anthropic’s Claude support page says the company is rolling out identity verification for some use cases, platform integrity checks, and safety or compliance measures, using Persona Identities as its verification partner. The page says verification data is used to confirm identity and not to train Anthropic’s models, but commenters focused on the privacy tradeoff, third-party data handling, and whether ID checks will restrict access to frontier models. The discussion also surfaced an important caveat: multiple users say this help page has been live since April, so the current outrage may be tied to broader anxieties about AI access controls rather than a brand-new announcement.
Discussion: Negative — The HN thread is overwhelmingly distrustful of Anthropic’s identity verification, especially the use of Persona and the possibility that ID checks could become a gatekeeper for advanced AI access. A smaller but important counter-thread notes that the support page and ID process appear to have existed for months, so some reactions may be conflating this with newer model-access controversies. (Privacy and biometric-data risk, Distrust of Persona as verification vendor, Fear of government or export-control-driven AI access tiers)
0:00 / 0:40policy Danish privacy activist says police raided him after posting PM’s numbers
Danish privacy activist and former police officer Lars Kragh Andersen says armed, masked police broke down his door after he obliquely posted Prime Minister Mette Frederiksen’s social security number and phone number, and shared a screenshot of a WhatsApp attempt to question her about encryption bans and surveillance powers. He claims officers immediately cut power to his router and removed Google Nest cameras, preventing the raid from being recorded, and says they refused to tell him the charges. The story matters because it sits at the collision point between doxxing, political protest, police accountability, and Europe’s ongoing fights over encryption and state surveillance.
Discussion: Mixed — HN was uneasy about the reported police tactics, especially cutting power and removing cameras, but commenters were sharply divided on Lars Kragh Andersen himself. Several Danish commenters described him as a provocative figure who has allegedly crossed lines with GPS tracking, family harassment, and doxxing, while others argued his stunts expose political hypocrisy around surveillance and privacy. (concern over police disabling recording and seizing cameras, debate over whether confrontational activism helps or harms a cause, claims of selective enforcement and hypocrisy by Danish authorities)
0:00 / 1:17general Free RTS Beyond All Reason Revives Total Annihilation at Massive Scale
Beyond All Reason is a free real-time strategy game inspired by Total Annihilation, emphasizing fully simulated units, projectiles, explosions, terrain deformation, and battles involving armies of thousands. The project drew attention on HN as a rare modern RTS with serious scale and polish, especially for fans of the TA and Spring RTS lineage. The comment thread, however, focused as much on the social experience as the game itself, with many users praising the design while warning that public multiplayer can be harsh for new or casual players.
Discussion: Mixed — HN commenters were highly impressed by Beyond All Reason as a technically ambitious, free RTS in the Total Annihilation lineage, with praise for scale, controls, performance, and nostalgia. But the dominant discussion quickly turned to multiplayer culture: several players described the community as toxic, meta-driven, and unfriendly to newcomers, while others argued the problem is mostly confined to popular 8v8 lobbies and can be avoided with solo, co-op, smaller games, or beginner-friendly rooms. (Strong nostalgia for Total Annihilation and 1990s RTS games, Praise for real-time simulation, scale, controls, and performance, Concerns about toxic multiplayer behavior and rigid meta expectations)
0:00 / 0:27general Google Says IPv6 Has Finally Hit Half the Internet
Google’s measurements show IPv6 crossing 50% of user access to Google services for the first time, a symbolic milestone for a protocol that has been rolling out for decades. APNIC cautions that its own global measurement is lower, at 42%, largely because it weights country-level samples by estimated Internet user populations rather than aggregating ad-delivered samples directly. The bigger point: IPv6 is now deployed at global scale, but adoption remains highly uneven by region, ISP, access network, and service provider.
Discussion: Mixed — HN treats the 50% mark as a real milestone, but the mood is more weary than celebratory. Commenters focus on lagging ISPs, IPv4-only major services, home-router pain, and the awkward incentives that keep dual-stack networks around, while also noting that IPv6 can already be faster and cleaner where CGNAT is the IPv4 alternative. (Frustration with ISPs that still do not offer residential IPv6, Measurement nuance: Google shows 50% while APNIC shows 42% due to different weighting methods, Major services and developer infrastructure still lacking IPv6 support)
0:00 / 0:26general Finland’s Libraries Are Lending Sewing Machines, Studios, and Social Glue
The BBC profiles Finland’s expanding vision of public libraries, especially Helsinki’s Oodi Library, where people can borrow or use sewing machines, 3D printers, music pods, sports gear, meeting rooms, and more. The article frames libraries less as book-lending institutions and more as democratic infrastructure: free public places for learning, making, social inclusion, and navigating modern bureaucracy. Finland spent nearly €371 million on public libraries in 2025, about €65.78 per person, and the Finnish Library Act explicitly tasks libraries with promoting democracy, freedom of expression, and active citizenship.
Discussion: Positive — HN readers were broadly enthusiastic, with many citing their own local “library of things” programs in the US, Canada, Europe, and Australia. The main caveats were practical: maintenance, waitlists, staffing, and the way underfunded libraries can be pulled into serving as de facto homeless shelters. (Libraries as civic infrastructure and third places, Borrowing infrequently used tools instead of buying them, Maker spaces, 3D printers, sewing machines, and media studios)
0:00 / 0:37general Hacker News marks Father’s Day with stories of dads, mentors, and making things
A “Tell HN” post wished the community a happy Father’s Day through a personal memory: the poster’s uncle, acting as a father figure in Soviet-controlled Poland, fostered a hacker-like childhood with rocket cars and an electrically functional police siren mounted to a banana-seat bike. The post resonated because it framed fatherhood broadly—as mentorship, curiosity, mischief, and the adults who help turn kids into makers.
Discussion: Positive — The thread is overwhelmingly warm and reflective, with many commenters sharing first Father’s Day stories, memories of technically curious dads, and gratitude for parents or father figures. The mood is not purely celebratory: several comments acknowledge difficult childhoods, infertility, grief, and the fact that Father’s Day is not universal or easy for everyone. (gratitude for fathers and father figures, new parenthood and early-childhood intensity, technical curiosity passed down through family)