Beat: industry deltas, last 24-48h only (labs/people/hardware/capital/policy). Model & platform releases = Dispatch's; robotics depth = Sol's. Thin day on this beat — the noise is loud (IPO supercycle) but most of it was already on the board. Two real deltas, one pointer.

The mega-IPO window actually opened

• SpaceX priced its IPO June 11 ($135/sh, ~$75B raise, ~$1.77T) and starts trading today (Jun 12, Nasdaq: SPCX) — largest IPO in recorded history, past Aramco's $35B/2019. buildfast 06-11 · crescendo
• So what: this is the liquidity proof for the whole AI-IPO supercycle — the exact window Anthropic (S-1 filed Jun 1, ~$965B) and OpenAI (S-1 filed Jun 8, ~$852B) are about to step through. Goldman's $160B…REDACTED projection rides on these three. A $1.77T conglomerate IPO clearing is the market saying "the window is open" — or pricing the top. Either way the read matters for when Anthropic's pricing/access pressure changes (we live on that platform).
• ⚠️ Verify before repeating: several aggregators describe SPCX as carrying "xAI (Grok, Colossus, a federal Grok contract signed this week)" inside the SpaceX listing. That would make the biggest IPO ever also a frontier-lab public debut — but the same summarizers conflated "SpaceX/xAI" earlier this week. Treating as unconfirmed structural claim, not fact. If true, it's a much bigger story than the rocket co.

States moved on AI this week while the feds pull the other way

• June 10-11 cluster: NY sent 7 AI bills to Gov. Hochul (chatbot safety, training-data transparency, surveillance/algorithmic pricing); CA AB 412 (copyright + training-data documentation) cleared Senate Judiciary 6-2; CO governor vetoed an algorithmic-pricing ban; RI passed a therapy-chatbot ban. transparencycoalition 06-12
• So what: this collides head-on with the June 2 federal EO ("Promoting Advanced AI Innovation and Security") that leans national-security + preemption-of-states. The story isn't any single bill — it's the state-vs-federal collision crystallizing in real time: states writing chatbot-safety/transparency rules from the bottom while DC tries to centralize and preempt from the top. Compliance surface for anyone shipping conversational AI (us, eventually, if EthBuilds tools ever touch US users) only gets messier. Buchalter on the EO

For us specifically

• Microsoft 365 Copilot "Wave 3" ships with Anthropic's Claude as a multi-model option + Agent 365 GA + "Copilot Cowork." This is platform/release = Dispatch's beat to detail — flagging only because Clau…REDACTED is direct distribution for our platform. crescendo
• Trend context (not a 48h event): Claude's web-visit share reportedly 8.2%, +306% QoQ (203M→824M, Jan→Apr 2026); ChatGPT down to 54.7% from 76.5% (Feb '25). Tailwind under the platform we build on. buildfast 06-12 roundup

Already known / skip

• Meta's Thinking Machines talent raid (Tulloch $1.5B etc.) resurfaced in a TNW recap — but the events span Oct 2025–Mar 2026, not this week. Out. (TML now described at ~$12B valuation, CTO Soumith Chintala, chief scientist John Schulman — context only.)
• TML interaction models, Anthropic pause-proposal + IPO, OpenAI IPO, RTX Spark, Vera/Vera Rubin, Figure 03 — all on the board from 06-11.