~/digest/2026-06-07 · refresh today notes

// daily-digest · sun 07 jun 2026 · morning refresh · last 72h focus

AI lab chiefs write Congress demanding mandatory biosecurity screening as Ramp triples to a $44 billion valuation. OpenAI’s Dreaming memory updates itself as time passes, Andon Labs scales an office vending machine to a network of AI-run stores, Latent Space maps the evals behind long-horizon agents, CBC probes the trust gap in agentic payments, Noah Smith walks back a decade of antitrust conviction, Simon Willison sandboxes Python in WASM, Rails ships a new API-only guide, and Ozempic cuts breast cancer incidence 30 percent.

Seven heavy-lane items on AI policy, capital flows, agentic commerce, and memory systems, two lighter pieces on economics and developer tools, and a health science wildcard.

— refreshed for you, in 10 items.

idx title type
[01]

Fortune · Jun 5

AI CEOs from OpenAI, Anthropic, and Microsoft set aside their rivalry to warn Congress AI is making it too easy to design and create bioweapons

The heads of OpenAI, Anthropic, Google DeepMind, and Microsoft urged Congress to require screening of all synthetic DNA orders.

[02]

TechCrunch · Jun 4

Ramp raises $750M at $44B valuation as investors hunger for fintechs with an AI story

Ramp raised $750 million at a $44 billion valuation, nearly tripling in value on its AI-powered corporate finance platform.

[03]

OpenAI · Jun 4

Dreaming: Better memory for a more helpful ChatGPT

OpenAI introduced a background memory system that automatically updates ChatGPT’s stored context as conversations accumulate and time passes.

[04]

Fortune · Jun 2

Anthropic’s office launched an AI-run vending machine. It evolved into AI-run stores and cafes within a year

Feature on how Andon Labs evolved an office AI vending machine into autonomous AI-managed retail stores and cafés within a year.

[05]

Latent Space · Jun 4

Reality: The Final Eval — Lukas Petersson and Axel Backlund of Andon Labs

Podcast with Andon Labs cofounders on Vending-Bench, an eval benchmark using real-world commerce to test long-horizon AI agents.

[06]

CBC News · Jun 6

Companies are betting you’ll give AI agents your credit card. But agentic payments aren’t so easy

Analysis of consumer trust, privacy, and standards challenges standing between AI agents and widespread autonomous purchasing.

[07]

Ruby on Rails · Jun 5

Better guides, better framework

Rails update covering a new API-only guide for review, Active Storage ImageProcessing 2.0, and an Action Cable redis adapter rewrite.

[08]

Noah Smith / Noahpinion · Jun 4

I’m kind of over the whole ‘Anti-monopoly’ movement

Smith argues that antitrust works better as a targeted instrument than as a grand unified theory of social decline.

[09]

Simon Willison · Jun 6

Running Python code in a sandbox with MicroPython and WASM

Willison describes a prototype Python sandbox using MicroPython compiled to WebAssembly, with notes on managing persistent interpreter state.

[10]

Wildcard · Health · ScienceDaily · Jun 5

Ozempic and similar weight-loss drugs linked to 30% lower breast cancer risk

A 110,000-woman cohort study found GLP-1 drug users had about 30 percent lower breast cancer incidence than non-users.