Sent to 20 readers
Race to AGI Daily Digest - Monday, June 8, 2026
Share:
TLDR
Google, SpaceX, and xAI lock in a $30B deal for 110,000 NVIDIA GPUs, tightening the frontier compute race.
Unisound’s 266B-parameter U2 agentic LLM shows how quickly tool-using, MoE-based systems are scaling.
AI chip stocks, led by ARM and peers, sold off sharply even as long-term GPU demand looks stronger than ever.
The Full Story
Last week we talked about Washington’s voluntary AI review scheme and the sense that control, not just capability, was becoming the real battleground. We also flagged the gap between calm public markets and a brewing fight over who gets access to frontier‑class compute.
Today that gap snaps into focus. On the tech side, China’s Unisound just rolled out its U2 agentic LLM, a 266B‑parameter Mixture‑of‑Experts model built to coordinate tools and services, not just chat. That is another data point in the quiet arms race around autonomous, multi‑step systems. If you want the specs, they’re here: Unisound U2 breakdown ->.
At the same time, Google has locked in a $30B deal with SpaceX and xAI for 110,000 NVIDIA GPUs. That is not just a cloud contract; it’s a statement about who gets first call on scarce hardware. You can dig into the numbers here: Google–SpaceX–xAI compute deal ->. If you want to zoom out on NVIDIA’s position in all of this, start here: NVIDIA profile ->.
Here’s the thing: while private deals chase more GPUs, public markets just hammered chip names. ARM, Intel, Qualcomm, AMD, and Baidu all fell around 10–13% today. So you have record demand for compute on one side, and a market suddenly asking if it overpaid for the AI hardware story on the other.
Meanwhile, last week we wondered whether US oversight talk would echo globally. That call is already showing up: China’s People’s Daily is warning about ‘AI pollution’ and data poisoning, and Deloitte is calling 2026 Australia’s year of white‑collar AI disruption. Both rhyme with the US narrative we’ve been tracking: US AI oversight tracker ->.
And out in the open web, engineers are asking if LLMs are quietly eroding their careers, while others just want a Linux desktop client from Anthropic. If you want to see where the jobs are actually moving, keep an eye on our live board: AI jobs map ->. This week is shaping up as a collision between bigger models, tighter oversight, and very human anxiety about what all of this means.
Get This Delivered Daily
Join thousands of AI professionals who start their day with Race to AGI.