Good morning. It's Tuesday, June 2, and we're covering an AI weather startup challenging the top forecasting agencies, the debate over whether AI is actually eliminating jobs, and Anthropic's quiet move toward IPO.

Plus: a reader poll on the Pope's AI manifesto, and a new piece from Google's Chief AI Business Strategist Gopi Kallayil.

Anthropic quietly files confidential IPO paperwork, giving itself a path to public markets without committing to a listing date. The AI company submits a draft S-1 registration statement to the U.S. Securities and Exchange Commission for a proposed initial public offering of common stock. Share count and pricing remain undisclosed, with any offering contingent on SEC review, market conditions, and other factors.

NVIDIA’s new DGX Station for Windows aims to run AI models with up to 1 trillion parameters locally, bringing data-center-class infrastructure to enterprise desktops. The system combines the GB300 Grace Blackwell Ultra Superchip with Windows-native management and security tools for AI development and agent deployment. It delivers up to 20 petaflops of FP4 performance, 748GB of coherent memory, and 800Gb/s networking. Microsoft says the platform extends Windows from traditional PCs to AI workstations built for always-on agents.

NVIDIA is pitching its new Vera processor as a CPU built specifically for AI agents, arguing that future data centers will optimize for tokens generated rather than traditional compute metrics. The company says Vera delivers 1.8x faster task completion than x86 processors for agentic AI, reinforcement learning, and data-processing workloads. The chip features 88 custom Olympus cores, up to 1.2TB/s memory bandwidth, and support from cloud providers including Oracle, CoreWeave, and ByteDance, with systems expected this fall.

MiniMax claims its new M3 model matches or exceeds leading proprietary systems while costing a fraction of the price, narrowing the gap between open and closed AI. The Chinese startup says M3 outperforms GPT-5.5 and Gemini 3.1 Pro on selected benchmarks, while offering a 1-million-token context window, multimodal capabilities, and strong coding performance. Promotional API pricing starts at $0.30 per million input tokens and $1.20 per million output tokens, rising to $0.60/$2.40 at standard rates.

ElevenAgents by ElevenLabs enables you to easily build intelligent, real-time conversational AI agents. By using your own knowledge base, SOPs and policies, ElevenAgents creates agents that sound like your brand in over 70 different languages at low latency and in minutes.

Ship voice agents for sales, support, and operations. Configure, deploy, and monitor natural, human-sounding agents that scale your team without scaling headcount.

Try ElevenAgents Free: CLICK HERE

The following Forward Future Original is by Gopi Kallayil, the Chief AI Business Strategist at Google.

The answers were surprising. First, it determined that guests cared more about breakfast than fine dining experiences. They’d been investing heavily in prestigious dinner restaurants with celebrity chefs, but guests really wanted an excellent breakfast. Second, 90% of breakfast orders were modified. 

So they made a bold decision: they removed the breakfast menu entirely. Through my work at Google and the years I’ve spent in Silicon Valley, I’ve witnessed multiple technological breakthroughs—from the internet to mobile phones to smartphones. And I’ve observed one consistent pattern that AI is now confirming: travel and hospitality are always among the first industries impacted by new technology. → Read the full article here.

Pope Leo XIV’s 40,000-word AI encyclical sparks debate on human dignity, AI risks, open-source innovation, and Anthropic’s growing influence.

WindBorne Systems has released WeatherMesh-6, an AI weather forecasting model that it says outperforms forecasts from the European Centre for Medium-Range Weather Forecasts (ECMWF), widely regarded as the global leader in weather prediction. The model generates hourly forecasts—rather than the standard six-hour cycle—and delivers resolution as fine as 3 kilometers across Europe and the continental U.S. WindBorne attributes the improvement to its network of roughly 400 weather balloons and advances in feeding real-time sensor data directly into its transformer-based AI system.

Founded in 2019, the Stanford-born startup has raised $25 million and currently sells weather data and forecasts to organizations including NOAA, the U.S. Air Force, and commodity traders. → Read the full article here.

A growing number of technology companies, including Meta, Coinbase, and Block, have linked recent layoffs to artificial intelligence, helping fuel concerns that AI is replacing workers. Yet a closer look suggests many cuts are also tied to broader business challenges, including overhiring during the pandemic, shifting corporate priorities, and pressure to fund expensive AI infrastructure.

More than 115,000 tech workers have been laid off across at least 150 companies in 2026, according to Layoffs.fyi, even as many firms report strong profits and continue hiring for AI-related roles. Economists say the clearest impact so far appears concentrated among junior technology workers and recent graduates, while executives increasingly frame restructuring efforts as preparation for an AI-driven future rather than simple cost-cutting. → Read the full article here. (Paywall)

Google Backs Sterile Mosquito Plan: Alphabet seeks approval to release 32 million sterile males to curb disease by shrinking mosquito populations.

SpaceX Flags Future Equity Issuance: New IPO disclosures warn of significant equity issuance in future transactions, fueling acquisition speculation.

Alphabet to Raise $80B for AI Compute: Alphabet will sell $80B in stock ($10B to Berkshire Hathaway) to fund its planned $180–$190B in 2026 AI capex.

Koji Launches AI Tutor: The startup says its AI tutor guides students through math and coding by prompting reasoning instead of giving answers.

Strava Tightens Data Access: Ahead of its IPO, Strava will lock more data behind logins, charge developers, and crack down on scraping.

AI Boom Sinks Older Unicorns: Nearly half of U.S. unicorns haven't raised funding in three years as AI reshapes valuations and investor priorities.

Thanks for reading. See you next time!

— Matthew Berman, Nick Wentz & the Forward Future Team