The name is the button.

An agent writes the code now, and a human thumb hovers over Accept all a hundred times a day. The honest question of this decade is what, exactly, you just accepted. So I read the diff.

Eight years as a backend engineer in Stockholm, back in Tokyo since 2023, and I still ship something small most nights, usually badly. My one rule: the only real review of a model is the thing you built with it. That is why there is no leaderboard on this site.

Around a hundred and forty sources come through — the labs' own posts, the papers, the tools, the runtimes and languages nobody is watching yet, the chip under the demo, and the newsletters people actually forward. What is moving gets counted so the week reads at a glance, and every few days I explain one mechanism in plain words.

A story here is a headline, a short excerpt, one image and the link to whoever wrote it. We never reproduce the post; the images belong to their authors. A post written in Japanese appears in Japanese, exactly as it was published, and the link goes to the original.

— Yuki Halvorsen, Tokyo

← Accept All Archive

SCMP TechAsia

As AI safety fears mount, can a US-China hotline prevent a global crisis?

September 21

The United States and China’s agreement to establish an official AI dialogue, including a proposed threat-notification system, marks a pragmatic step towards crisis prevention in their tech war, even…

Regulation & safety

Read at SCMP Tech ↗

One letter a week, in your inbox.

Latest

As AI safety fears mount, can a US-China hotline prevent a global crisis?

Related

AsiaChinese regulator drafts rules to protect teenagers from ‘intimate’ AI companions SCMP Tech

SiliconAnthropic, OpenAI, SpaceXAI, and Google face antitrust lawsuit for agreeing to slow AI development Tom's Hardware AI

LabsIntroducing the Australian Youth Safety Blueprint OpenAI News

VoicesLiability, regulation, and AI’s new false dichotomy Gary Marcus

VoicesHow a single tweet transformed the AI safety debate Understanding AI

ValleyLLMs respond differently to harmful prompts when AI watermarking is used Ars Technica AI