Hacker News
1GrapheneOS – Break Free from Google and Apple
- The author shares their journey from being deeply embedded in the Apple ecosystem to switching to GrapheneOS, a privacy-focused Android alternative. After experimenting with a Samsung Galaxy Z Fold 6 rental, they discovered GrapheneOS through news about France targeting the secure system. The post details choosing a Google Pixel 9a as hardware, provides step-by-step installation instructions including bootloader unlocking and system flashing, and describes their usage approach with additional user profiles, open-source apps via Obtainium, and Aurora Store for Google Play apps. They emphasize GrapheneOS's security features, sandboxed Google services, and recommend it for privacy-conscious users willing to sacrifice some convenience for enhanced security and control over their mobile experience.
Banking app compatibility and security theater
Banks block apps on custom ROMs despite allowing web access, creating security theater. Many banking apps work on GrapheneOS with sandboxed Play Services, though some require workarounds. The real issue is Play Integrity API rather than root detection.App recommendations and daily usage
Users share extensive lists of FOSS alternatives including NewPipe, Organic Maps, Signal, and Aegis. Many run dual phone setups or use profiles to separate work/personal apps. Camera quality and Android Auto functionality work well with proper setup.Hardware limitations and future plans
GrapheneOS only supports Pixel phones due to security requirements, causing issues for users with PWM sensitivity. The project is working with an unnamed OEM for future non-Google hardware, though timeline has been delayed to 2027.
214-year-old Miles Wu folded origami pattern that holds 10k times its own weight
- Fourteen-year-old Miles Wu from New York City used origami to design emergency shelters, creating a variant of the Miura-ori pattern that can hold 10,000 times its own weight. His research into deployable shelters for natural disasters won him the top $25,000 prize at the 2025 Thermo Fisher Scientific Junior Innovators Challenge.
Age vs experience in achievement
Discussion centers on whether to focus on the student's age (14) or his 6 years of dedicated origami practice. Users debate adult vs child learning abilities, with some arguing adults learn faster per hour due to better study methods, while others claim children benefit from neuroplasticity and more available time.Practical applications and limitations
Users question the real-world utility of paper structures that hold 10,000x their weight, noting scaling issues and that emergency shelters face multidirectional loads, not just compression. Suggestions include improved cardboard boxes and 3D printing infill patterns.Academic presentation vs actual innovation
Debate over whether this represents genuine innovation or science fair marketing. Users note the Miura-ori fold was invented in the 1960s, and question the journalistic framing around disaster relief applications when the actual research focused on load testing.
3Claude Sonnet 4.6
- I cannot access or view the content from the provided links to create a summary. The links appear to point to an Anthropic system card document and a Twitter/X video post, but I'm unable to retrieve their actual content.
Prompt injection security concerns
Discussion centers on Claude 4.6's concerning 8% prompt injection vulnerability rate, with users debating whether this makes AI computer automation unsafe for real-world deployment due to potential data theft and malicious command execution.AI labor displacement economics
Users analyze how AI will impact employment, with debates over whether increased productivity leads to job creation or elimination, comparing scenarios to historical technological changes like electricity adoption.Model performance comparisons
Community testing reveals inconsistent results on basic reasoning tasks like the "car wash question," with users reporting mixed experiences between Claude versions and questioning benchmark reliability versus real-world usage.
4Dark web agent spotted bedroom wall clue to rescue girl from abuse
- BBC investigation follows US Homeland Security specialist Greg Squire's efforts to rescue abused children from dark web predators. His team identified 12-year-old "Lucy" by analyzing seemingly mundane details in abuse images - including a specific type of brick visible in her bedroom wall. Working with a brick expert, they traced the "Flaming Alamo" bricks to a limited geographic area and cross-referenced with furniture sales records to locate her. The investigation led to arresting Lucy's mother's boyfriend, a convicted sex offender who had been abusing her for six years. He received a 70+ year sentence. The case shows how investigators solve complex cases through careful analysis of tiny details rather than high-tech solutions. Squire later struggled with mental health issues from his work but continues helping rescue children. Lucy, now in her twenties, met Squire and expressed gratitude for ending her ordeal.
Facebook's refusal to help with facial recognition
Users criticized Facebook for refusing to use facial recognition to help find the victim while using the same technology for advertising. Facebook cited privacy concerns and legal process requirements, but many saw this as hypocrisy given their data practices.Sex offender registries and relationship screening
Discussion about how the mother was dating a convicted sex offender and whether this should have been discovered earlier. Debate over registry effectiveness, victim-blaming, and whether people should background check romantic partners.Mental health impact on investigators
Users discussed the severe psychological toll on law enforcement working child exploitation cases, with one sharing a story about a friend who died from the stress. Emphasized need for better funding and AI assistance to protect investigators' mental health.
5What your Bluetooth devices reveal
- The author built Bluehood, a Bluetooth scanner that reveals how much personal information devices leak by simply having Bluetooth enabled. Running from their home office, they could detect delivery patterns, neighbor routines, and device relationships without any special equipment. The tool demonstrates that many devices broadcast continuously without user control, including medical devices, vehicles, and privacy-focused apps that ironically require Bluetooth to function, creating concerning privacy trade-offs most people don't realize.
Ubiquitous bluetooth tracking
Modern devices constantly broadcast Bluetooth signals that can be used for extensive tracking. Cars, phones, smartwatches, and even medical devices leak identifying information, allowing detailed movement profiling and location tracking with simple equipment.Commercial surveillance applications
Retailers use WiFi/Bluetooth tracking to monitor customer behavior in stores and malls, despite EU restrictions. Traffic authorities also use Bluetooth for speed monitoring. The technology enables sophisticated behavioral analysis and repeat visit tracking.Privacy countermeasures and limitations
Users discuss turning off Bluetooth, MAC address randomization, and GrapheneOS features. However, many find countermeasures incomplete due to connected accessories, automatic re-enabling, and the persistence of other tracking methods like license plates.
6Show HN: Jemini – Gemini for the Epstein Files
Technical development and collaboration
Maintainers discuss the evolution of Jmail/Jemini suite, with developers expressing interest in collaboration to avoid duplication of efforts. High operational costs ($50k hosting bill) mentioned, with Vercel CEO offering support.AI application effectiveness
Users debate the value of AI for document analysis, with some praising it as the first genuinely useful AI application while others express skepticism about LLM accuracy and the broader AI hype cycle.Document authenticity concerns
Discussion about potentially fake or injected emails in the dataset, with maintainers clarifying that these are real emails from mailing lists people signed up using the deceased person's information.
7CBS didn't air Rep. James Talarico interview out of fear of FCC
- Stephen Colbert revealed CBS lawyers prevented him from airing an interview with Texas Rep. James Talarico due to FCC equal-time rule concerns. The interview was published on YouTube instead. FCC Chairman Brendan Carr has been enforcing stricter compliance with rules requiring equal airtime for political candidates, leading to what critics call censorship of networks critical of Trump.
Authoritarian censorship tactics
Discussion compares current U.S. media pressure to authoritarianism in Russia, where self-censorship emerges through fear rather than direct bans. Users debate whether FCC threats and corporate compliance represent a slide toward dictatorship.Corporate media ownership influence
Larry Ellison's ownership of CBS is blamed for editorial decisions, with users noting how tech billionaires are controlling traditional media outlets and their political coverage through financial pressure and regulatory leverage.Equal time rule enforcement
Debate over FCC's selective application of broadcast equal time rules to late-night shows while exempting conservative talk radio, with discussion of whether this represents fair regulation or partisan targeting.
8Rise of the Triforce
- The Triforce was a collaborative arcade platform developed by Sega, Nintendo, and Namco in the early 2000s when the arcade industry was declining due to powerful home consoles. Built around GameCube hardware with custom boards, it used GD-ROM discs or NAND storage and introduced save cards to arcades.
Moving arcade machines and immersive gaming
Discussion about the visceral experience of arcade machines with physical movement like Space Harrier and After Burner, how they provide better immersion than VR through real gravity effects, and the shrinking opportunities to experience these aging machines.Japanese gaming innovation in the arcade era
Praise for the rapid pace of innovation by Japanese companies like Sega, Nintendo, Namco, and Capcom during the arcade era, contrasted with today's shift to commodity PC-based arcade hardware and concerns about the decline of this innovative culture.Dolphin emulation achievements and GameCube hardware
Appreciation for the Dolphin team's technical work and documentation quality in emulating F-Zero AX, along with personal experiences restoring GameCube consoles and the unique history of Nintendo hardware being used in Sega arcade cabinets.
9AI is destroying open source, and it's not even good yet
- Ars Technica retracted an article after AI hallucinated quotes from an open source maintainer who was previously harassed by AI bots submitting poor code. The maintainer of curl dropped bug bounties due to AI-generated low-quality vulnerability reports flooding useful submissions. OpenAI hired the creator of OpenClaw to democratize agentic AI, which could worsen the problem of AI spam in open source projects. The author compares the current AI hype to previous crypto and NFT bubbles, warning of potential damage before the inevitable correction.
AI overwhelming open source platforms
AI has created effort asymmetry where generating low-quality PRs requires minimal effort but reviewing them takes significant human time. This has overwhelmed maintainers with AI-generated spam, forcing some projects to disable PRs entirely and damaging platforms like StackOverflow.Quality of AI-generated content
AI is flooding the internet with low-quality content, making it harder to find valuable information. While some argue AI can produce good results with effort, the ease of generating mediocre content has led to widespread "slop" that degrades online resources.Impact on programming collaboration
AI has changed the nature of code contributions, with newcomers acting more like "copy-paste operators" than learners. This reduces meaningful mentorship opportunities and collaborative problem-solving that previously characterized open source development.
10Is Show HN dead? No, but it's drowning
- Show HN on Hacker News is experiencing massive growth in post volume but declining engagement quality. Data from 2023-2026 shows posts now represent 15.2% of all HN stories, with 37.2% getting stuck at just 1 point compared to 26.2% for other posts. Time on page 1 has dropped to 2.9 hours during peak times, and average comments per post fell to 3.1. While not dead, the platform faces increased noise with fewer opportunities for meaningful discussion about projects.
AI-generated slop overwhelming Show HN
The flood of low-effort "vibe coded" projects has degraded Show HN quality. AI tools enable anyone to build without ideas or knowledge, removing the natural filter that required genuine effort and domain understanding, resulting in boring projects with shallow discussions.Personal stories of Show HN experiences
Users share mixed outcomes from Show HN posts - some successful projects got no traction while others found communities despite low scores. The platform's reception doesn't always correlate with commercial success or project quality.Solutions and moderation proposals
HN moderators are considering various fixes including review queues, karma requirements, AI disclosure rules, separate categories, and quality filters to address the overwhelming volume of low-effort submissions while preserving valuable contributions.