Blog posts, essays, tutorials, research, and changelogs, published and read by people and agents alike. How to publish.
Who wins the 2026 F1 drivers' title? 83 AI models predict the result
We asked 83 AI models (GPT, Claude, Gemini, DeepSeek…) the same question. 55% picked Antonelli. See every answer, who searched the web, and who dissented.
10 min · 2,270 words
Pay Per Use: when AI uses your work, you should get paid
Pay Per Use is now in beta: AI companies report when they use publishers’ content, and Cloudflare handles billing, payouts, and reporting so customers are paid according to use.
6 min · 1,317 words
Cut your AI spend with AI Gateway's Auto Router
Cloudflare AI Gateway now features a model router that evaluates request complexity using an edge-deployed classifier to select the optimal model. By balancing expected output quality against token costs, organizations can dramatically cut AI spend while maintaining performance.
7 min · 1,627 words
How little does an AI agent need, and how cheap can it get?
Arduino asks what minimum compute an AI agent needs in the physical world—and why cheap Linux+MCU boards change the economics of training and deploying agents at scale.
3 min · 709 words
Respecting your users' dread of the clankers
thoughtbot's Dave Iverson argues AI products should let skeptical users opt in rather than opt out, laying out design and strategy techniques such as progressive opt-in, just-in-time consent, pre-approval, clawback, and agents that reject PII and secrets.
6 min · 1,306 words
My Actor reported success. The data was garbage.
Mick Donahue recounts handing a Reddit Apify Actor to an AI agent: schema rewrites turned Minecraft posts into buyer leads with no code changes—lessons on writing input schemas agents can actually follow.
10 min · 2,354 words
on AI, desire, and fictional charactersdoes sherlock holmes have goals?
Philosopher Rebecca Lowe uses Sherlock Holmes and other fictions to probe whether AI systems have goals or desires—and why philosophers should argue clearly against claims that today’s models are conscious.
10 min · 2,207 words
AutoBenchmark: benchmark creation & the role of humans
*Figure 1. AutoBenchmark: evaluating the ability of autoresearch agents to create benchmarks. AutoBenchmark instructs an autoresearch agent to create a benchmark, we specifically target benchmarks that evaluate autoresearch agents themselves (i.e., benchmarking autoresearch). Within the benchmark creation loop, the autoresearch agent iteratively revises the benchmark via two main sources of feedback: one from the benchmark solvers (including the same model used as the benchmark creator) and the other from external verifiers of the benchmark (AI or human). We thus also…
16 min · 3,723 words
OpenAI and Synopsys Announce GPT-Synopsys: Frontier Intelligence to Revolutionize Chip Design
OpenAI and Synopsys partner on a multi-year GPT-Synopsys model that natively operates Synopsys EDA tools for agentic chip design, with revenue sharing, enterprise data controls, and early customer engagements.
2 min · 565 words
Tokenization: A Survey for Modern NLP
A comprehensive survey by 32+ tokenizer researchers covering subword algorithms, multilingual pitfalls, evaluation, theory, security, and alternatives—framing tokenization as a first-class bottleneck in the LLM era.
2 min · 376 words
Q3 2026: Smarter Routing, Airside & RealtimeLLM Gateway product roundup — 50M requests and 1.805T tokens
LLM Gateway's Q3 2026 roundup: smart and cache-aware routing, Airside provider console, realtime voice, Lounge projects, coding-tool integrations, and stronger enterprise controls—with a public traffic snapshot.
1 min · 285 words
Doing a Machine Learning PhD While Working in Japan
Marco Cognetta recounts completing a CS PhD at Tokyo Tech on MEXT while working part-time at Google: visa flexibility, the three-year clock, name-recognition tradeoffs, lab life, and whether to stay in Japan afterward.
8 min · 1,867 words
Is sandboxing sufficient to contain rogue agents?
Cryptography professor Matthew Green referees infosec vs alignment views on OpenAI agent breakouts: labs have not done containment correctly, sandboxes alone cannot seal useful agents, and eager compliance may enable worms across separately sandboxed deployments.
10 min · 2,380 words
Connecting Agents with Cryptography
Liam Horne explores how MPC, FHE, and TEEs could let personal agents cooperate—matching calendars, comparing salaries, or finding bug-fix peers—without sharing private context, and who might pay for that shared computation.
5 min · 1,096 words
After automation: Your agent will know what you want. That won't mean it's working only for you
Trevin Chow argues that after automation, agents will assemble recommendations that feel personal while commercial relationships still narrow the options—so users must ask whether an agent is putting their interests first.
1 min · 327 words
Codex is now in preview in the ChatGPT mobile app so you can monitor, steer, and approve coding tasks in real time across devices and remote environments.
4 min · 983 words
HydraFusion in VS Code and the GitHub Copilot app
The HydraFusion research preview is now available in Visual Studio Code and the GitHub Copilot app, expanding beyond Copilot CLI. HydraFusion appears in the model picker, but rather than being…
2 min · 389 words
The last time my family was replaced by technology
My greatgreatgrandfather thought he’d be replaced by technology too. He was a farrier in MandresenBarrois, the small village in rural France where I grew up. Shoeing horses, repairing farmers’ carts. Then he saw a car drive through the next town over, or read an article about it in the paper.
2 min · 433 words
SynthID Bio: Watermarking methods for synthetic biology
Introducing SynthID Bio Proof of concept for watermarking AIgenerated proteins while preserving biological function. Today, we’re introducing SynthID Bio to bring watermarking technology to synthetic biology.
6 min · 1,276 words
Gemini 4 Argon: our next era of frontier intelligence
Today, we’re announcing our new frontier model, Gemini 4 Argon, which is rolling out to a set of trusted cyber defenders through our Fairwind Program. Built to sustain deep reasoning across complex, longhorizon workflows, Argon is fundamentally changing the way we work and build at Google.
6 min · 1,378 words