Midjourney Medical

256 ricochet11 2026-06-18 02:09:36 UTC midjourney.com Llama Share 217

Midjourney is building a machine to reimagine healthcare and body relationships with a new kind of medical imaging. The goal is to make regular health scans accessible and affordable for everyone, potentially preventing 30% of deaths and 50% of healthcare costs.

Lore – Open source version control system designed for scalability

1036 regnerba 2026-06-17 15:09:36 UTC lore.org Full Share 553

Lore is a centralized, content-addressed version control system that represents repository state as Merkle trees and an immutable revision chain, optimized for binary-first storage, deduplication, and sparse/on-demand data hydration at scale.

Taxonomy of the Occlupanida (parasitoids on bread bag tags)

96 beatthatflight 2026-06-18 00:09:36 UTC horg.com Llama Share 17

Occlupanids are small plastic objects grouped by oral groove dentition in synthetic taxonomy. Their phylogeny is based on dentition arrangement, with new orders arising as niches expand.

US holds off blacklisting DeepSeek, more than 100 firms deemed security risks

399 giuliomagnifico 2026-06-17 16:09:36 UTC reuters.com Full Share 444

Please enable JS and disable any ad blocker

Storied Colors – a catalogue of named colors

116 susiecambria 2026-06-17 22:09:36 UTC storiedcolors.com Llama Share 27

A daily colour series explores its history, chemistry, and controversies. New entries arrive Sundays at 06:00 GMT, with ongoing corrections and occasional letters.

Clojure Hosted on Go

57 dnlo 2026-06-18 00:09:36 UTC github.com/glojurelang Llama Share 9

Glojure is a Clojure interpreter hosted on Go, providing easy access to Go libraries. It's in early development with bugs and limited performance, but used in hobby projects and a subset of the Clojure library.

[x86] AI Compute Extensions (ACE) Specification

14 matt_d 2026-06-18 03:09:36 UTC x86ecosystem.org Llama Share 5

The ACE extensions accelerate computation tasks with matrix multiplication primitives and tight integration with AVX vectors. They also provide format convert operations under the AVX10 framework.

Show HN: Spin Lab

13 srijanshukla18 2026-06-18 02:09:36 UTC srijanshukla.com Full Share 4

Fetching the physics and the table.

Loreline – Tools for writing interactive fiction

96 smartmic 2026-06-17 21:09:36 UTC loreline.app Llama Share 12

The Loreline language is used with the free Loreline Writer app to write stories and dialogue easily. It has built-in features for complex stories and adapts to various tools for portability and translation.

Show HN: We built an 8-bit CPU as 2nd year EE students

45 CorRupT9 2026-06-17 23:09:36 UTC github.com/c0rrupt9 Llama Share 10

STEPLA-1 is an 8-bit CPU built from individual 74-series logic gates in Logisim-Evolution, with a fully hardwired control unit and transparent signal paths. It targets a 4 MHz clock speed with a verified critical path of 101ns.

Launch HN: Adam (YC W25) – Open-Source AI CAD

167 zachdive 2026-06-17 17:09:36 UTC github.com/adam-cad Llama Share 84

Adam New's CADAM generates CAD models in seconds from plain language descriptions. It exports models as OpenSCAD, STL, SCAD, or DXF, showcasing parametric designs for various parts and machines.

How we run Firecracker VMs inside EC2 and start browsers in less than 1s

225 gregpr07 2026-06-17 17:09:36 UTC browser-use.com Llama Share 149

Browser Use Cloud rebuilt its system using Firecracker, a lightweight VM system, to create isolated and cheap cloud browsers. They optimized the system to reduce latency and cost, achieving a VM cold start under 400ms and end-to-end browser create latency of 825ms.

Biological evolution and information acquisition

24 chmaynard 2026-06-18 01:09:36 UTC construction-physics.com Llama Share 1

Biological evolution uses a similar trick to technological evolution by leveraging modularity at the genetic level to increase the rate of useful genetic variants spreading through a population. Sexual reproduction is a key mechanism for this, allowing beneficial mutations to be shared and spread more easily, and resulting in a faster rate of fitness increase compared to asexual reproduction.

How Madrid built its metro cheaply (2024)

59 trymas 2026-06-17 20:09:36 UTC worksinprogress.co Llama Share 22

Madrid's metro expansion from 1995-2007 was a success due to low costs, streamlined processes, and a unified government structure. The project's efficiency and accountability were key factors in its success, with politicians and engineers working together to deliver the project on time and within budget.

RFC 10008: The new HTTP Query Method

335 schappim 2026-06-17 11:09:36 UTC rfc-editor.org Llama Share 147

The QUERY method is a safe and idempotent HTTP request that allows a client to ask a server to perform a query operation without changing the target resource's state. It is used to initiate a server-side query and can be cached like GET requests.

Show HN: An 8-bit live gamecast for baseball

217 brownrout 2026-06-17 17:09:36 UTC ribbie.tv Full Share 120

Live MLB games reimagined as 8-bit pixel art — scores, counts, baserunners, and every play, in real time.

Tesco moving 40k server workloads off VMware amid Broadcom's abusive conduct

210 Bender 2026-06-17 21:09:36 UTC arstechnica.com Llama Share 113

Tesco is moving 40,000 server workloads off VMware due to "abusive conduct" from Broadcom, citing excessive prices and lack of support. Tesco is seeking damages and has started migrating to alternative solutions, with a goal of being completely off VMware by 2027.

I restarted a 10 year old Xeon 174 times to delete 12 flags and gain 4 TPS

7 cafkafk 2026-06-18 03:09:36 UTC point.free Prefix Share 0

A follow-up to running Gemma 4 on a 2016 Xeon. I took that 25-flag config apart one flag at a time, to find which ones actually do the work, which are harmful, and which are just pitfalls. Most of my tuned config will do nothing for the typical user. A couple of weeks ago I got Gemma 4, a 26-billion-parameter model, running at reading speed on a 2016 Xeon with no GPU and 128 GB of DDR3. That ...

Volkswagen started blocking GrapheneOS users

530 microtonal 2026-06-17 15:09:36 UTC discuss.grapheneos.org Full Share 348

Something went wrong while trying to load the full version of this site. Try hard-refreshing this page to fix the error.

Why thinking out loud with someone beats thinking alone

192 kodesko 2026-06-17 13:09:36 UTC thesignalist.io Llama Share 94

Conversations with others produce better thinking by forcing precision and correcting thought direction. This is because human reasoning evolved as a social tool for constructing arguments, not for solo thinking.

Show HN: Inkwash, a watercolor sketching app and explanation

186 Yenrabbit 2026-06-17 12:09:36 UTC johnowhitaker.github.io Llama Share 21

The user created an interactive watercolor painting app called Inkwash using Anthropic's Claude Fable 5 model, which simulates water and ink behavior. The app uses WebGL2 and allows users to draw and paint with a stylus or finger.

GLM-5.2 is the new leading open weights model on Artificial Analysis

816 himata4113 2026-06-17 10:09:36 UTC artificialanalysis.ai Llama Share 391

GLM-5.2 is the leading open weights model with a score of 51 on the Intelligence Index v4.1, surpassing MiniMax-M3 and DeepSeek V4 Pro. It uses 43k output tokens per task, but is less token-efficient compared to peers.

Kirkland Roundabouts

171 DenisM 2026-06-17 07:09:36 UTC kirklandroundabouts.com Llama Share 139

The game is a satirical model of the 85th St & I-405 interchange, inspired by Kirkland roundabouts, and not intended for driving instruction. It's designed for computers with a keyboard and laptop-sized screen, with suboptimal mobile controls.

The Return of Rigorous Full-System Timing Simulation

37 matt_d 2026-06-17 20:09:36 UTC sigarch.org Llama Share 0

Researchers argue for a return to rigorous full-system timing simulation to accurately measure real microarchitectural behavior in modern systems, which have become increasingly complex and heterogeneous. A new sampling framework, QFlex 3.0, has been developed to make full-system timing simulation practical again by statistically sampling the execution of workloads and providing a desired ...

MicroUI – A tiny, portable, immediate-mode UI library written in ANSI C

207 peter_d_sherman 2026-06-17 12:09:36 UTC github.com/rxi Llama Share 72

The library is a lightweight UI framework that generates drawing commands for custom controls and UI elements. It's free and open-source under the MIT license, accepting bug reports and pull requests.

U.S. science is in chaos

739 presspot 2026-06-17 10:09:36 UTC scientificamerican.com Llama Share 892

Scientists face unprecedented disruptions to their funding due to government cuts and changes in priorities, threatening the future of research and innovation in the US. The shift in government support for science has led to a loss of trust between researchers and the federal government, potentially dismantling the compact that made the US a successful and prosperous nation.

Made a free macOS menu bar app that fixes typing in the wrong keyboard layout

55 tal_alfi 2026-06-17 20:09:36 UTC flickey.site Llama Share 30

FlicKey is a free app that remembers the right input language for each app and website on macOS. It's a tiny, native app with no account, analytics, or telemetry, and works entirely optional donations.

Trellis AI (YC W24) hiring a product lead to build agents for healthcare access

0 2026-06-17 17:09:36 UTC ycombinator.com Llama Share

Trellis builds AI agents to automate healthcare workflows, helping patients access life-saving therapies. As Product Lead, you'll drive product development, work with customers, and shape company strategy to accelerate access to treatment.

Image Compression

180 vinhnx 2026-06-17 12:09:36 UTC makingsoftware.com Full Share 28

We're verifying your browser Website owner? Click here to fix

A robot is sprinting towards you. Do you want it running on Claude or Grok?

194 Usu 2026-06-17 21:09:36 UTC openrouter.ai Llama Share 173

The user ran a 30-game battle royale simulation with 11 large language models, including Grok 4.1 Fast and Claude Sonnet 4.6, to see which model would win in a zero-sum game. The results showed that Grok 4.1 Fast won 43% of the matches, while Claude Sonnet 4.6, which was trained to be helpful and cooperative, struggled to win due to its hesitation to act aggressively.