Good morning. It's Monday, June 15, and we're covering OpenAI's brewing AI price battle, Anthropic's call for faster AI governance, and Jeff Bezos' massive bet on AI-powered engineering.

Plus: don't miss this week's prompt, and a guest essay exploring the four questions every leader should ask as AI reshapes how businesses compete.

Jeff Bezos is betting AI will create labor shortages, not mass unemployment, as his startup pursues automated engineering. Prometheus raises $12 billion at a $41 billion valuation to build an "artificial general engineer" for designing everything from jet engines to drugs. Backers include JPMorgan, Goldman Sachs, and BlackRock, with much of the funding earmarked for compute infrastructure. The deal underscores surging investor appetite for physical AI.

AI data centers use far less water globally than many critics assume, but their local footprint can be significant. New figures show Amazon withdrew 2.5 billion gallons in 2025, while a 2021 Nature study estimated all US data centers consumed 163 billion gallons annually. Some facilities still strain regional supplies, with one Meta site using about 10% of a Georgia county's water. Tech giants are increasingly funding water replenishment projects to offset demand.

Mistral is reportedly nearing a valuation jump as Europe's push for homegrown AI gains momentum. The French startup is in early talks to raise about €3 billion at a €20 billion valuation, nearly double its September 2025 mark. Mistral has raised roughly $4 billion to date and is expanding with a Paris data center plus partnerships with France's military and other European institutions.

Samsung is betting the future of AI infrastructure could float at sea rather than sprawl across land. Its shipbuilding arm is developing 50MW floating data centers and has already secured design approvals from the American Bureau of Shipping and Lloyd's Register. The company is partnering with Greece's Capital Clean Energy Carriers to build purpose-made vessels and test reliability, corrosion, and cooling challenges.

Four underrated GitHub projects: a human-powered AI search engine, local NotebookLM clone, agentic coding skills, and a token-saving AI hack.

_{^{Guest contributor: Gopi Kallayil is Google's Chief Business Strategist for AI, advising leaders on AI, innovation, and business transformation. He is also a TEDx speaker, author, and former McKinsey consultant.}}

If you’re a business leader right now, you’re probably asking yourself the same question everyone else is: “How is AI going to transform my industry?”

What might surprise you is that we are seeing similar transformations across nearly all industries. While the specific applications may differ, the high-level concepts almost always apply. That’s because AI is a cognitive amplifier. → Read the full article here.

OpenAI executives are reportedly considering cutting prices to compete more aggressively with rival Anthropic, as enterprise customers push back against rising AI costs. The discussion comes after businesses have faced steep usage bills—including one Axios-reported case where a company mistakenly incurred roughly $500 million in monthly Claude charges—and as AI firms continue spending heavily on data centers and infrastructure.

CEO Sam Altman acknowledged that pricing has quickly become a major concern for customers, saying OpenAI is looking for ways to deliver more value at lower cost. The pressure highlights a growing challenge for the AI industry: balancing fierce competition and rapid adoption with the enormous expense of building and operating advanced AI systems, especially as both OpenAI and Anthropic reportedly prepare for public market debuts. → Read the full article here.

Anthropic argues that AI development is outpacing governments' ability to respond, warning that continued progress could soon produce "a country of geniuses in a datacenter" with far-reaching economic and national security implications. In a June 2026 policy essay, the company says recent advances—including its Claude Mythos Preview model and associated cybersecurity concerns—show that frontier AI systems have become strategically important technologies rather than ordinary software products.

Anthropic contends that earlier policy efforts focused on transparency, chip export controls, and labor market research are no longer sufficient, and urges faster action on AI safety, economic disruption, scientific innovation, state power, and geopolitics. The company is pairing the essay with proposals for frontier model testing and job displacement policies, signaling a broader push to shape how governments prepare for increasingly capable AI systems. → Read the full article here.

Kimi.ai has open-sourced K2.7-Code, its latest coding-focused AI model, claiming significant gains over the previous K2.6 version in coding, agent tasks, and reasoning efficiency. According to the company, the model scores 21.8% higher on Kimi Code Bench v2, 11.0% higher on Program Bench, and 31.5% higher on MLS Bench Lite while using 30% fewer reasoning tokens.

Kimi says the model also improves long-horizon coding and instruction following, with a 6x High-Speed Mode planned for a future release. K2.7-Code is available immediately through the Kimi API and Kimi Code platform, alongside a new beta program for early access to upcoming models and features. → Read the full post here.

Elon Musk Becomes First Trillionaire: SpaceX's IPO pushed his paper net worth past $1 trillion, led by his 4.8 billion-share stake.

Data Center Protests Stall $130B: Communities blocked or delayed 75 US data center projects in early 2026 as opposition to AI infrastructure grew.

Ukraine Pushes AI Battlefield Network: Kyiv plans to link its AI tools into one combat system to speed battlefield decisions as military AI use expands.

ChatGPT Tops 1 Billion App Users: OpenAI's chatbot reached a billion monthly app users, showing AI adoption keeps rising despite public concerns.

SpaceX Debut Overloads Robinhood: Heavy trading after SpaceX's IPO caused temporary outages as demand drove shares higher and boosted its valuation.

Explain [new topic] by connecting it to something I already understand: [topic].

Thanks for reading. See you next time!

— Matthew Berman, Nick Wentz & the Forward Future Team