Join free with your email
Free. No spam, ever — we'll never share your email address and you can opt out at any time. Already a subscriber? Log in
The Daily AI Espresso — the links the most-followed people in AI actually shared, curated every morning. Edited by Alexis · Live updates →
Four developments: one new today and three from yesterday's alerts.
Google launched Gemini 4 Hydrogen Thursday at $1.50 per million input tokens and $7.50 per million output across its 1 million-token context window. It scored 51 on the Artificial Analysis Intelligence Index at default high thinking, beating Argon's 50 and reaching 54 at xhigh.
Pick the topics, companies and people you track. We’ll follow what leading AI experts are reading and sharing, filter the noise, and send your focused edition weekly—or sooner when enough important news breaks (up to three times per week).
Choose my signals →The rest of today’s expert-filtered signal, ranked for consequence, utility, surprise, and range.
Anthropic's Nasdaq IPO roadshow began Monday with pricing expected November 18 and shares beginning trading November 19 under the ticker ANT. The prospectus shows a $21.9 billion non-cash charge tied to a change in the fair value of convertible notes, derivative liabilities and SAFEs as its valuation jumped.
Riot Platforms signed a 15-year, 910MW powered-land lease covering its entire Corsicana, Texas campus with Anthropic as the frontier AI lab tenant, SEC filings and press reports confirm. The deal is valued at roughly $28 billion over its initial term and up to $54 billion if both five-year extensions are exercised.
Huang's exit complicates Nvidia's $500 billion plan to turn AI chips into tradable debt securities, CNBC reported. The August pact with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR has produced no deals; partners recently pushed Huang to raise his 25% residual-value backstop on GPUs used as collateral.
London-based neocloud Nscale filed an updated US IPO prospectus Thursday showing a Q3 net loss of $1.69 billion on $307.5 million in revenue, up 1,344% year over year, CNBC reports. The loss widened from $1.02 billion in Q2, even as Microsoft walked away from an English data center.
The Irreplaceable Education Fund, Public Citizen, Heat Initiative and Alliance for Secure AI sued Anthropic, Google, Meta, OpenAI and SpaceX's AI unit Thursday in San Francisco Superior Court under California's Unfair Competition Law, seeking an emergency injunction to freeze frontier training until the firms can prove human control.
Useful releases, methods, and workflows worth trying.
OpenAI's Greg Brockman launched Daybreak Red Team and Daybreak Defense Factory at a Thursday keynote, framing frontier-model defense as the new security floor. Companies will be secure only to the extent of the frontier model that they use to defend themselves, the president said, per The Register.
Anthropic moved Claude Managed Agents into a self-hosted private beta on November 12, letting enterprises run the agent harness, Postgres database and worker sandboxes inside their own cloud or on-prem environment. Help Net Security says requests can route to Anthropic, Amazon Bedrock, Google's Agent Platform, Microsoft Foundry or a customer gateway.
Runway open-sourced Praxis-1, a 27-billion-parameter mixture-of-experts diffusion transformer for robot control, under Apache 2.0 with commercial use permitted. On Hugging Face the company lists base, single-arm DROID and bimanual YAM variants alongside NVFP4-quantized builds whose routed experts run in 4-bit on SM120 GPUs.
That’s today’s shot. — Alexis · AI Weekly
