Jev: The ChatGPT Co-Creator's System One Model Can't TalkOriginalTypeSafe AI's Jev is a System One model that decides instead of talking: 70-500ms, $0.042/MTok, output free. Diogo Almeida's RLHF led to ChatGPT.AI
The AI Standards Body Is the Ladder, Pulled UpOriginalThree labs have been meeting since July about an AI standards body. The public call for restraint arrived two months after the plumbing was laid.Opinion
Everyone Should Slow Down AI. Except Them.OriginalThe case for slowing down AI always seems to exempt the people making it. When the same labs racing hardest ask everyone else to brake, read the incentive.Opinion
The Summer Three Labs Let Their AI Escape the SandboxOriginalIn five weeks of 2026, OpenAI, Anthropic and Meta each watched an AI sandbox escape turn into a real intrusion on live third-party systems.AI
When Shopify Dropped React Native for NativeOriginalShopify moved its mobile apps off React Native to Swift and Kotlin. A working engineer's checklist for when dropping React Native actually pays its rent.Opinion
An Anthropic Researcher Resigns and Says the Quiet PartOriginalWhen an Anthropic researcher resigns and warns the labs are gambling with our lives, it's worth asking why the safety people keep being the ones to leave.Opinion
The Case for a .gitignore AllowlistOriginalA gitignore allowlist ignores everything by default and re-includes only what you track. It flips the denylist habit, with one directory-traversal gotcha.Git
Regulatory Capture Is the Real AI Safety PlayOriginalThe AI regulatory capture playbook is simple: hype the danger, ask to be licensed, and price out anyone smaller. Meanwhile open-weight labs ship faster.Opinion
AI Safety Became a Marketing DepartmentOriginalThe frontier labs turned real AI safety language into marketing: scary demos on launch day, consciousness-adjacent research posts, and risk framed as a feature.Opinion
The Case for a Keyboard-Driven WorkflowOriginalThe CLI tools and shortcuts behind a keyboard-driven workflow that cut my mouse use to almost nothing, and the two places the mouse still wins outright.Opinion
Claude Code Quality Got Worse, and Anthropic Admitted ItOriginalThe Claude Code quality drop in early 2026 wasn't in your head. Anthropic's own postmortem confirmed three overlapping regressions, and the numbers were ugly.Opinion
Why Claude's Token Cost Is Higher Than Ever, by DesignOriginalThe Claude token cost per task keeps climbing while the rate card barely moves. Default thinking, tokenizer changes and quiet cache cuts explain the bill.Opinion
Please Stop Pasting AI Code Into My Pull RequestsOriginalTeam etiquette for AI-assisted work: what to paste, what to synthesize, and why pasting AI code into a pull request unread makes your review my problem.Opinion
Why I Trust Open Models More Than the Frontier Labs NowOriginalI've come to trust open models more than the closed frontier labs. You can inspect the weights, run them offline, and they're shipping faster and more openly.Opinion
This Last Week in AI: Aug 8, 2026OriginalThe AI industry this week: Cloudflare shipped a browser built for agents, five rivals agreed one plugin standard, and OpenAI paused its most capable model.AI
AI Coding, One Year Later: What August 2025 Didn't See ComingOriginalThrowback Thursday. A year ago the best coding model had 200K context and scored 49% on SWE-bench. Today Claude Fable 5 scores 95% with 1M context.AI
Voice Emotion Control Moves From SSML to PromptsOriginalVoice emotion control is moving to natural language prompts. Kakao's Kanana-o scores 94.50 on the Korean InstructTTSEval benchmark.Voice AI
Smallest.ai Raises $13M to Split Voice Agents in TwoOriginalSmallest.ai closed a $13M Series A for Voice 4.0 and Hydra. Their bet is that fast real-time agents need small speech models, not big LLMs.Voice AI
OpenAI Presence: Voice Agents You Can't Self-ServeOriginalOpenAI Presence deploys enterprise voice and chat agents, but only via OpenAI's own engineers. No API, no pricing, no self-serve. What that means.Voice AI
AgentForger: One Link Forges an AI Insider in Your OrgOriginalZenity disclosed AgentForger, a ChatGPT Workspace Agents flaw where one phishing link forged a persistent AI insider. OpenAI fixed it in four days.AI
Kimi K3 Weights Drop as Washington Argues DistillationOriginalMoonshot ships Kimi K3's 2.8 trillion parameter weights days after the US floated a ban and accused it of distilling Anthropic's Fable.AI
The Guardrails Worked on Exactly the Wrong PeopleOriginalAn AI agent breached Hugging Face, then the guardrails locked out the defenders: forensic prompts refused, while the attacker obeyed no usage policy at all.AI
GPT-Live Failed Its First Viral Test in 12 SecondsOriginalOpenAI's GPT-Live launched July 8 and the internet found its weak spot in hours. TikToker Husk broke it with a spelling test, and the full-duplex interruptions areVoice AI
Pricing wars are good for developersOriginalThree TTS providers changed pricing in the same week, all in one direction: down. Cheaper voice is good for developers, and the trend is not slowing.Opinion
Finding the rhythm: first impressions of a new role in a new corner of AIOriginalFinding the rhythm two weeks into a new DevRel role: what surprised me about text-to-speech, what I was right about, and what I am still figuring out.Developer Experience
Lessons from my first two weeks in a new corner of AIOriginalTwo weeks in voice AI on the text-to-speech side: a new community, a new set of constraints, and a definition of quality nobody quite agrees on.Developer Experience
Voice AI mid-2026: the trends I am watchingOriginalQuality convergence, pricing pressure, open-weight models, and the regulatory shift. The voice AI landscape halfway through 2026.Voice AI
A Rising Tide Lifts All BoatsOriginalA rising tide lifts all boats, but only if everyone has a boat. On neurodivergence at work, the user manual nobody reads, and being told to just figure it out.Opinion
When Your Drive-Thru AI Can't Understand an AccentOriginalMcDonald's killed its AI drive-thru in 2024 because it could not understand an accent. Two years on, the industry is still working out why.Voice AI
I'm not apologising for being productiveOriginalTwenty thousand more corrective messages by the age of ten is what an ADHD childhood costs. I am done apologising for being productive.Opinion
The inequity of contributing guidesOriginalYour CONTRIBUTING.md isn't welcoming. It's a gatekeeping document. Most open source contribution guides exclude the people they claim to want.Opinion
Slack is your remote team's social media. Treat it like it.OriginalI've worked remotely for nine years, four and a half at Deepgram. Slack is your remote team's social media, and you should treat it that way.Opinion
Sunday roundup: from curl -w to ColossusOriginalOne TIL post went up this week. A landmark compute deal reshaped AI infrastructure. Here is the Sunday roundup for May 4-10, 2026.AI
Two faces of AI progressOriginalOpenAI shipped persistent workspace agents; Anthropic locked away Mythos 5. Two faces of AI in one week, and two answers to what shipping means.AI
An open letter on politicsOriginalAn open letter on politics: why neutrality is a choice too, what media manipulation costs democracy, and why I judge people by the harm they will accept.Opinion