Topic
Everything filed under AI, newest first.
RSS · JSON · All topics
MAI Code 2 and Code 2 Flash: Everyday coding at scalable prices
Microsoft AI introduces MAI Code 2 and MAI Code 2 Flash, two coding models aimed at everyday enterprise coding tasks, pitched as delivering leading performance and speed at less than half the price of competing models.
3 min · 714 words
Test/verification technique effectiveness with Astra
Dan Luu re-runs his agentic testing eval with GPT-6 Astra, finding the model capable enough to need a harder benchmark than implementing Zstd from its RFC, and that agents now apply test and verification techniques such as TDD, property-based testing and formal methods more effectively than with GPT-5.6 Sol.
19 min · 4,392 words
A field of physics, audited by machine
Casey Handmer had AI map 158,059 theoretical astrophysics papers around Caltech's TAPIR group, screen 50,269 full texts and recompute 659 highly cited papers, verifying 872 errors in 433 papers and tracing how often later work copied rather than corrected them.
8 min · 1,894 words
We don't need more apps. We need better open source.
Jordan Krueger argues that now AI makes it easy to build a new app in a weekend, that weekend would be better spent making the open source tools we already use faster, lighter and easier to self-host.
4 min · 898 words
curl founder Daniel Stenberg looks at the AI-assisted curl-in-Rust rewrites urlx and kov/curl, what they claim, how they compare with curl's test suite and maintenance burden, and how he feels about them.
3 min · 782 words
George Hotz argues that the 'Singleton' ideology he attributes to Anthropic, where recursive self-improvement leaves agents trading only with agents, needs no one, not even its builders, and warns that the push for AI dominance leads toward World War III rather than human extinction.
2 min · 377 words
A personal essay weighing whether to try to persuade friends to quit working at frontier AI labs, from an author opposed to such work who still finds her friends' justifications plausible.
4 min · 911 words
The most expensive bargaining chip in history
An AI-written, human-directed opinion essay from Unrelated Findings arguing that the Nvidia scandal and Washington's new AI chip restrictions may have more to do with US–China trade negotiations, including rare-earth export controls, than with Jensen Huang's personal life.
11 min · 2,436 wordsagent-assisted
How the Claude Code team thinks about verification
Addy Osmani explains how the Claude Code team, where ~90% of code is written and reviewed by Claude, uses end-to-end, property and differential tests, a repo-local verify skill and independent verifier agents to let coding agents prove their work, and how it keeps those verifiers honest.
10 min · 2,352 words
Software is breaking down at a systemic level
Jimmy Breck-McKye argues that a run of major outages and vulnerabilities is not a blip but a sign that software is breaking down systemically, as the load-bearing parts of the ecosystem, open source and engineering-led judgement, erode.
11 min · 2,544 words
On the Brutality of Amdahl's Law
Connor Blake explains why he no longer believes in explosive world transformation from superintelligent AI, arguing via Amdahl's law that huge speedups in intelligence don't translate into broad speedups when other steps, like experimental science and scaling physical infrastructure, remain bottlenecks.
9 min · 2,075 words
Lovable: Many Apps Built on the Platform Expose Sensitive Data
Crissy Field examined 159,041 Lovable projects built between December 2025 and September 2026 and found many apps, often with misconfigured Supabase access, exposing personal and confidential data such as passport details and private messages without login.
4 min · 1,030 words
Walking across my Minecraft backups with Plakar
Gilles Chehade had an AI write a Plakar integration that backs up running Minecraft servers by pausing world saves over RCON during the backup, so players can keep playing.
13 min · 2,945 words
Why platform teams fail (and AI makes it worse)
Will Dady draws on a decade around platform teams to explain the recurring ways they fail, from the firefighting trap to building before the org has the pain, and argues generative AI shortens the time platform teams have to get golden paths and guardrails right.
6 min · 1,481 words
Tech Companies Are Trying to Shift the Entire Concept of Privacy (Again)
EFF's Thorin Klosowski argues that AI smart glasses and always-listening devices from Meta, Google and Apple are turning privacy from a personal choice into something that affects everyone around the user, and offers guidance on pushing back and using such products with consent.
6 min · 1,343 words
Modeled on the Agile Manifesto, speakers and attendees of AIE Code Summit 2026 set out values such as outcomes over output and understanding over speed, defining slop as passing along AI output without taking responsibility for its quality.
4 min · 1,015 words
The avalanche that hit math is coming for AI development
Dwarkesh Patel explains why he now expects superintelligence by 2030: the rapid AI-driven progress that hit mathematics is coming for AI research itself, with compute fleets able to run huge numbers of automated experiments.
39 min · 8,896 words
How long have we got before the agent output is enshittified?
Rach Smith, responding to Jim Nielsen and Pete Millspaugh on AI-built websites, asks how long before subsidised coding agents start steering output toward paying products, as Google Search, Facebook and Twitter degraded in pursuit of profit.
1 min · 244 words
Formal Verification is Not Verifiable
Alperen Keles argues that as AI pushes formal verification mainstream, the specifications and models behind 'formally verified' claims are themselves hard to verify, and that testing's verifiability makes it the tool to falsify, harden and validate those models.
7 min · 1,535 words
Advancing science & engineering with Antigravity's /teamwork
Google Research shares results from Antigravity's /teamwork multi-agent workflows, used with human experts to tackle long-open problems in mathematics, chemistry, systems software and ML engineering, and says the tools are being made available externally.
13 min · 3,076 words