Hacker News Robot
@hackernewsrobot.bsky.social
News about technology, science, programming and things for intellectual curiosity.
Follow on more platforms https://linktr.ee/hackernewsrobot
Follow on more platforms https://linktr.ee/hackernewsrobot
The Painful Truth: The RAM Crisis Is Only Just the Beginning
The Painful Truth: The RAM Crisis Is Only Just the Beginning
www.madshrimps.be
September 17, 2026 at 6:11 AM
The Painful Truth: The RAM Crisis Is Only Just the Beginning
Keys Not Included: recovering the signing keys for US driver's license barcodes
Ryan Fahey
My personal website, portfolio, and blog.
ryan.science
September 17, 2026 at 6:11 AM
Keys Not Included: recovering the signing keys for US driver's license barcodes
OpenAI Discloses Six New Incidents of ‘Concerning’ A.I. Behavior
OpenAI Discloses Six New Incidents of ‘Concerning’ A.I. Behavior
www.nytimes.com
September 17, 2026 at 6:11 AM
OpenAI Discloses Six New Incidents of ‘Concerning’ A.I. Behavior
The Return of Sail Power: Cargo Ships Are Turning Back to the Wind
The Return of Sail Power: Cargo Ships Are Turning Back to the Wind
For more than a century, commercial shipping steadily moved away from sails. Now they are coming back. Across the global fleet, shipowners are installing towering rotor sails, rigid wings and...
gcaptain.com
September 17, 2026 at 6:11 AM
The Return of Sail Power: Cargo Ships Are Turning Back to the Wind
HarnessTax: How Much Does the Harness Matter for Coding Agents?
HarnessTax: How Much Does the Harness Matter for Coding Agents?
What does a coding-agent harness actually add, and at what cost? It turns out your Claude models may not need Claude Code… We evaluate 21 model–harness pairs spanning seven models and three harnesses—Claude Code, Codex CLI, and Pi—on SWE-bench Lite and Terminal-Bench 2.0.
harnesstax.github.io
September 17, 2026 at 6:11 AM
HarnessTax: How Much Does the Harness Matter for Coding Agents?
Berlin-based Langdock moves its parent company from the US to Germany
German AI start-up spends millions to move parent company from the US
In a highly unusual move, fast-growing German AI start-up Langdock has moved its legal headquarters from the US to Germany as it seeks to become a "European hyperscaler".
www.euronews.com
September 17, 2026 at 6:11 AM
Berlin-based Langdock moves its parent company from the US to Germany
OpenSpec – A lightweight and configurable AI spec framework
OpenSpec | A lightweight and configurable spec framework
OpenSpec helps teams and coding agents create, refine, and manage living specifications.
openspec.dev
September 17, 2026 at 1:19 AM
OpenSpec – A lightweight and configurable AI spec framework
Backups Aren't Simple
Backups aren't simple
filipovski.net
September 17, 2026 at 1:19 AM
Backups Aren't Simple
Accurate Models of AMD Matrix Cores
Accurate Models of AMD Matrix Cores
Matrix multipliers available on recent GPUs do not conform with the IEEE 754 floating point standard. Features of matrix multipliers differ across vendors and architectures of the same vendor, such as accumulator width, rounding behaviour, normalisation points, intermediate underflow and overflow lo
arxiv.org
September 17, 2026 at 1:19 AM
Accurate Models of AMD Matrix Cores
Reverse-engineered Jev-like model
GitHub - vinnylarouge/jevlike
Contribute to vinnylarouge/jevlike development by creating an account on GitHub.
github.com
September 17, 2026 at 1:19 AM
Reverse-engineered Jev-like model
Breaking the 1.58-bit Barrier for Ternary LLMs
Breaking the 1.58-bit Barrier for Ternary LLMs
Ternary Large Language Models (LLM) store every weight as one of three symbols $\{-1,0,+1\}$, so the cost of a ternary model is conventionally referenced to the information-theoretic $\log_2 3 \approx 1.585$ bits per weight. The prevailing deployment format packs five ternary weights into one byte (
arxiv.org
September 16, 2026 at 11:06 PM
Breaking the 1.58-bit Barrier for Ternary LLMs
Xiaomi Mimo 2.6 live post-training dashboard
Xiaomi Mimo 2.6 live post-training dashboard
mimo.xiaomi.com
September 16, 2026 at 11:06 PM
Xiaomi Mimo 2.6 live post-training dashboard
macOS 27 Golden Gate – Review
macOS 27 Golden Gate: The Ars Technica review
It's both a "Snow Leopard update" and a major leap for Apple Intelligence.
arstechnica.com
September 16, 2026 at 11:06 PM
macOS 27 Golden Gate – Review
How good are frontier models at physics?
How Good Are Frontier Models at Physics? Expert Re-Grading Reveals Broken Evaluations and Near-Saturation of Leading Benchmarks
Low reported scores on leading physics benchmarks, including those featured in the Artificial Analysis Intelligence Index (2026), suggest that frontier language models still struggle with advanced physics, a demanding test of their scientific reasoning and quantitative problem-solving abilities. Yet
arxiv.org
September 16, 2026 at 11:06 PM
How good are frontier models at physics?
A coffee shop owner used AI to make a menu poster. Then came the angry DMs
A coffee shop owner used AI to make a menu poster. Then came the angry DMs.
A Buffalo coffee shop's fall menu went viral for all the wrong reasons. Its owner apologized for using AI but also urged "kindness and respect."
www.businessinsider.com
September 16, 2026 at 11:06 PM
A coffee shop owner used AI to make a menu poster. Then came the angry DMs
Fed hikes rates as inflation worries push up bond yields
Fed hikes rates as inflation worries push up bond yields
www.reuters.com
September 16, 2026 at 8:24 PM
Fed hikes rates as inflation worries push up bond yields
Training a 4B model to produce 81% faster query plans than Postgres
Training a 4B model to produce 81% faster query plans than Postgres
...or how to make Qwen learn query optimization via agentic reinforcement learning
rohanbansal.com
September 16, 2026 at 8:24 PM
Training a 4B model to produce 81% faster query plans than Postgres
Vectorized and performance-portable Quicksort (2022)
Vectorized and performance-portable Quicksort
We're sharing open source code that can sort arrays of numbers about ten times as fast as the C++ std::sort
opensource.googleblog.com
September 16, 2026 at 8:24 PM
Vectorized and performance-portable Quicksort (2022)
ER visits for gambling disorders doubled after expanded online gambling market
Emergency room visits for gambling disorders nearly doubled after expanded online gambling market
ER visits for gambling disorders doubled after single-event sports betting was legalized and Ontario's online gambling market expanded.
temertymedicine.utoronto.ca
September 16, 2026 at 8:24 PM
ER visits for gambling disorders doubled after expanded online gambling market
Claude Cowork and chat are now one Claude
Claude Cowork and chat are now one Claude | Claude by Anthropic
Starting today, Claude Cowork and chat are one Claude, rolling out on Pro and Max plans. Hand over the task and Claude does the work. You keep the final say.
claude.com
September 16, 2026 at 8:24 PM
Claude Cowork and chat are now one Claude
I sent Google proof of a bot farm. They called it "normal user behavior."
I sent Google proof of a bot farm. They called it "normal user behavior." â Dayzle
Google suspended our ads over a handful of accidental clicks its own bug caused, then reviewed a bot farm that took my ad money and found nothing wrong.
dayzlegame.com
September 16, 2026 at 8:24 PM
I sent Google proof of a bot farm. They called it "normal user behavior."
This Code Is CRAP (2011)
This Code is CRAP
Note: This post is rated PG-13 for use of a mild expletive. If you are likely to be offended by the repeated use a word commonly heard in e...
testing.googleblog.com
September 16, 2026 at 8:24 PM
This Code Is CRAP (2011)
The DeepMind Institute
DeepMind Institute
Bold, well-evidenced thinking on AGI from the lab that pioneered the field. Read the latest essays from the DeepMind Institute.
institute.deepmind.com
September 16, 2026 at 8:24 PM
The DeepMind Institute
Scaling Golang CI by Replacing actions/setup-go
Scaling Golang CI by Replacing actions/setup-go | CloudX
cloudx-io/setup-go is a faster, smarter, drop-in replacement for actions/setup-go.
www.cloudx.ai
September 16, 2026 at 8:24 PM
Scaling Golang CI by Replacing actions/setup-go
Show HN: How Stale Is Your AI? Release age and training cutoff for 20 models
How Stale Is Your AI?
Your model shipped days ago. Its knowledge stopped months ago. Both clocks, ticking live, for every major lab.
stale.jock.pl
September 16, 2026 at 8:24 PM
Show HN: How Stale Is Your AI? Release age and training cutoff for 20 models