1741
Lean Eval for Alignment on Faithfulness
We turn intent into mathematics, verify every step formally, and make sure the machine does what it is supposed to.
New
The latest blog posts to take off on Hacker News. 5,947 posts.
1742
How we used to get jobs: A newspaper classifieds story
A Newspaper Classifieds Story
1743
CSS properties you should know for better text designs
Includes background-clip for masking backgrounds, vertical-align for aligning elements, box-decoration-mode for consistent edge styling, letter-spacing for spacing control, and text-combine-upright for vertical text layouts.
1744
Go is an ideal language for AI-assisted software engineering
As AI shifts software engineering from writing to reviewing, discover how Go's strict compiler and unified toolchain ensure reliable AI-generated code.
1745
Mojo 1.0
Today, the Mojo language officially reaches 1.0: a milestone the language has been building toward since its first release in 2023. Mojo has grown into a general-purpose language with a vibrant developer community writing their own libraries, tools, and applications on top of it. With Mojo 1.0, developers can now…
1746
Manus will return to operating as an independent company
Visit the Manus blog to learn more about our thoughts on the future.
1747
What I learned by putting GitHub Copilot behind a MitM proxy
A look inside Copilot’s network traffic, harness, memory, and how context is becoming the product.
1748
Nvidia's Risky Business
Nvidia is finding new ways for its customers to raise money, and it’s expanding the risk of the AI buildout significantly.
1749
1750
Don't classify, hallucinate
A useful pattern for LLM classification at scale is to let it hallucinate plausible, fake entities. Then resolve to real ones.
1751
Stop Killing Games: It's time to sue Sony, join us
Heb jij een PlayStation? Dan betaal je onterecht een Sony Tax op al je digitale games. Stichting Massaschade & Consument voert collectieve actie voor een Fair PlayStation.
1752
Rust SIMD on the GPU
At VectorWare , we are building the first GPU-native software company . Today, we are excited to announce that we can successfully use Rust's portable SIMD ( core::simd ) on the GPU. This milestone marks a significant step towards our vision of enabling developers to write complex, high-performance applications that…
1753
Learning more about Claude's mathematical capabilities
An unreleased Claude model improved the lower bound for the fraction of zeroes of the Riemann zeta function that satisfy the hypothesis, raising it from 41.6% to 67.2%.
1754
GPT 5.6 Cyber
Meet GPT-5.6-Cyber, OpenAI’s cybersecurity-specific model available through Daybreak Red for authorized vulnerability research, exploit validation, and security testing.
1755
What's the best programming language for coding agents?
This somewhat widely cited post (I keep seeing it cited, anyway) suggests that dynamic languages and/or languages that represent things more concisely are more token efficient. It seems to be cited enough that LLM search results agree. For example, when I searched for "dynamic vs static language token cost" (no…
1756
Itadakimasu: A word you say to the food, not the cook
On a small Japanese word that everyone has heard and almost everyone has misaddressed, and what changes when you find out who it is for.
1757
A fortuitous decade as an indie software developer
Ten years and one day ago, on August 9, 2016, I abruptly quit my job. I mark this occasion as the beginning of my indie software developer career. Happy anniversary to me and to my customers!
1758
Felix and I
He is young – twenty four – sitting at the black computer table by the wall of his living room. I’m looking at him side-on. He’s silhouetted by the huge window with a gorgeous view of the neighbourhood church. Green trees outside. The sun is shining brightly. It’s summer. Probably “The Hot Summer”, when our friends G…
1759
Letter to Governor Abbott on responsible AI infrastructure in Texas
OpenAI sent Governor Greg Abbott a letter outlining its commitment to responsible AI infrastructure in Texas. The letter supports reliable, transparent growth that benefits Texans.
1760
Exploring Claude/GPT Knowledge Cutoffs and Pre-Training Timelines
A mildly interesting analysis of what models know and what it tells us about how they were trained.
1761
The lattice of sets of natural numbers is rich (2021)
$\newcommand\of{\subseteq}\newcommand\N{\mathbb{N}}\newcommand\R{\mathbb{R}}\def\<#1>{\left\langle#1\right\rangle}\newcommand\Z{\mathbb{Z}}\newcommand\Q{\mathbb{Q}}\newcommand\ltomega{{{<}…
1762
Humanising LLM Outputs Is Dumb
The largest tell for me to tell where culture and sentiment is shifting for AI tools is usually X, viral GitHub repositories and Hacker News. One of these tells I’ve been seeing a lot lately is skills like I have ADHD and Agents.md instructions such as giving outputs in only ASD-STE100 Simplified Technical English.
1763
Defending my own brain against enshittification
If you haven't had Belvita breakfast cookies, you have probably had Biscoff cookies, which are similar. These are a classic airline snack. You've probably ha...
1764
Asynchronous I/O in DuckDB: Work, Thread, Work
Starting with v2.0, scheduled for fall 2026, DuckDB will support asynchronous reads of Parquet and CSV files. This can significantly speed up queries when synchronous I/O does not saturate the available bandwidth, as is typical in EC2/S3 compute-storage setups.
1765
Tl;dv: Over 180k meetings left wide open
How a missing Firestore security rule on tl;dv exposed 181,874 meetings from 84,312 users across 35,003 domains, including live calls I could join uninvited, and how six months of disclosure got me nothing but seen receipts.
1766
Squeak 6.1
These release notes are optimized for viewing inside Squeak, as they contain a lot of interactive examples. On this page, interactive links open in SqueakJS, which is a browser-based Smalltalk VM with some limitations. For the best experience, download Squeak and read the release notes there.
1767
Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows
Muse Glimmer is a 30-billion-parameter open agentic model from Meta Superintelligence Labs, optimized for always-on local workflows on consumer hardware.
1768
Auto mode is now the default in Claude Code
Claude Code will soon run auto mode by default for Pro, Max, Ultra, and Team plans, enabling longer-running autonomous work, and catching more dangerous commands.
1769
What Happened to HackerOne?
The rise and fall of the largest bug bounty platform in the world
1770
How We Pushed CDC into Postgres
Discover Snowflake Postgres Mirroring. Learn how push-based change data capture delivers transactional replication to your data lake with zero lag.