The Gemini app is now on Mac
Google is bringing the Gemini app to macOS as a native desktop experience.
Every blog post that took off on Hacker News, ranked by the votes of its best submission. 2,243 posts.
Google is bringing the Gemini app to macOS as a native desktop experience.
The following is part of a series of posts about 2026 summer intern projects – for more, see “What the interns have wrought, special jumbo 2026 edition”
Fixes 23 issues (addressing 13 👍). Run multiple scripts in --parallel or --sequential. bun:test mock & spyOn get Symbol.dispose support. ESM bytecode compilation. Faster Bun.markdown. Faster RegExp, String.prototype.trim, String.prototype.startsWith.
Back in the late 1970’s and early 1980’s Sandia National Laboratory (in Albuquerque NM USA) began building the capacity to design, fab, and test IC’s at scale (packaging was handled by Fairchild and Allied Signal). Why would a National Laboratory need the capacity to do this? To provide components that were not…
If you've ever clicked on a search result only to land on a version of a site you don't prefer, Kagi's URL Redirects (also known as URL Rewrites) puts you in control by automatically rewriting search...
The trademark infringement issue has now been resolved. The author of the website and project in question has removed all uses of the Notepad++ trademark from his product, website, and related materials. The unauthorized references have been taken down, and the trademark infringement is no longer ongoing.
How Go 1.26's source-level inliner works, and how it can help you with self-service API migrations.
NanoGPT Slowrun is an open effort to implement data-efficient learning algorithms; 5.5x data efficiency in the first week and improving.
Optimisation of bloom filters in Floe using double bit setting with lightweight hashing.
Time to upgrade
An introduction to diffusion language models and the research advances that underlie today's diffusion LLMs. We describe the building blocks of recent open-source models, starting from simple masking diffusion, and including techniques for iterative refinement, post-training, and variable-length generation. Material…
Anthropologist Roberto González (Professor, San José State University) examines how America’s military-industrial complex has been rapidly expanding from the Capital Beltway to Silicon Valley. Although much of the Pentagon’s budget is spent on conventional weapons systems, the Defense Department has increasingly…
Airbus is working at full throttle to offer the German Air Force an operational Uncrewed Collaborative Combat Aircraft (UCCA) system by 2029.
Forking pg_query.rs to replace Protobuf with bindgen-generated direct C-to-Rust struct mappings, yielding 5x faster query parsing and 10x faster deparsing in PgDog.
We asked five AI assistants 200 questions that call for a number and followed 1,285 of the statistics they cited back to the source.
Starting 2027, Google will block apps on Android devices whose developers have not registered with Google, paid, and given their ID. It’s fair to say that Google is becoming Apple: Your device, their rules.
This guide is written for practitioners wanting to create disk images of floppy disks with the intention of preserving them for the long-term. This guide will focus on 8-inch, 5.25-inch, 3.5-inch and 3-inch floppy disks and will only focus on getting material from these disks and will not cover writing disks. Other…
After frontier AI systems came up empty, AISLE surfaced six CVEs in curl, one of the world's most audited codebases. Its maintainers patched all six.
How a statistical approach cut IPFS content publishing times by over one order of magnitude.
Somewhere at the bottom of the Forth and Clyde Canal, buried in the silt, there is a box of custom silicon that was right about almost everything. Back in 1988 a Scottish hi-fi company shipped a processor that checked the type and bounds of every memory access in hardware, garbage-collected its own heap in silicon, […]
Model training has grown complex enough to require many specialised stages and teams, and manual coordination between them doesn’t scale. At Aleph Alpha, we’ve built Savanna, a model factory that implements the entire training pipeline in code.
Cunyan Ma’s interactive tool exposes the power structure that extracts low-wage labor from data workers in the Global South to make everyday AI algorithms function.
Two architectures for running agent harnesses, the tradeoffs between them, and how we make skills and memories work when the harness isn't local.
We are announcing a new Crystal release 1.22.0 with several new features and bug fixes.
A configuration mismatch in GLM 5.3 on Workers AI triggered an SGLang bug that exposed KV cache pages across requests. Here is how we addressed the issue.
Fragment of the Hillsborough meteorite, broken on impact, with fusion crust from passing at high speed through the Earth’s atmosphere.
Almost every traditional database keeps its write-ahead log and data files on one machine's disk, which is the root cause of data loss risk, expensive read replicas and high-availability clones, and analytics queries that drag down transactions.
More than 93,000 visible appearances of brands promoting unhealthy food and drink, alcohol, gambling and speculative products were recorded during coverage of the largest FIFA World Cup in history, new research has found. Unhealthy food and drink brands accounted for more than seven in ten, with 65,722 appearances.
How I burned a full Claude limit in 30 minutes and built my own deep research pipeline instead: 3 subscriptions, shared memory, a clear role for each model. You can build the same from what you already pay for.