Tonight's
Rundown.

Roughly 200 grads booed and walked out on Sundar Pichai over Project Nimbus — the $1.2B Google + Amazon contract with the Israeli military — and Google's ICE ties. Pichai pointedly never said the word "AI."

VC Vinod Khosla called the students "idiotic, short-sighted and very selfish"; Rep. Ro Khanna defended their right to protest. Google fired 28 employees in 2024 over the same contract.

The Malaysia-based AI messaging platform closed a $62.5M Series B led by Camber Partners (Endeavor Catalyst in too) — up from a $7M Series A in 2022 — at $35M ARR and 169% year-over-year growth.

"Per conversation" pricing means AI replacing human agents grows revenue instead of shrinking it. The platform runs ~2 billion messages a quarter for clients like Toyota, British Airways and Hertz — the basis of its data-flywheel "moat" pitch.

SpaceX priced 555.5M shares at $135, raising $75 billion — the largest IPO on record — then surged 19% to close at $161, valuing it around $2.1 trillion and making Musk the world's first trillionaire.

The rocket business is funding an AI money pit: the xAI division posted a $6.36B operating loss in 2025, and roughly $600–700B of the valuation is essentially a pure bet on xAI and orbital computing — both losing money today.

Same compute crunch as your lead story: Anthropic's president has said compute costs are pushing it toward its own IPO, and it confidentially filed its S-1 this month.

Rhode Island's House and Senate both approved a therapy-chatbot ban (H 7349) — part of a wider state wave: New York moved to ban AI chatbot toys, and Colorado's governor vetoed a bill that would've prohibited algorithmic pricing.

States are racing ahead of Washington. A New York law taking effect Nov. 5 will require companion chatbots to disclose they aren't human and to route users expressing suicidal thoughts to the 988 crisis line.

Access vs. safety: bots offer 24/7, low-cost mental-health support in underserved areas — against real risks to wellbeing and privacy when there's no clinician oversight.

A paper by Anthropic's Marina Favaro and co-founder Jack Clark argues AI systems are approaching the ability to recursively self-improve — and that humans are losing the ability to meaningfully oversee the process.

A globally coordinated, verifiable pause — not a unilateral halt — explicitly because if only one company stopped, competitors would race ahead.