🕶️ OpenAI and Anthropic share 90 VCs

Portfolio theory has finally eaten venture conviction. Ninety firms backing both labs in the most important category of the decade tells you all you need to know about which side capital is on, its own.

Meanwhile, compute contracts hit monthly rent most seed rounds would kill for, islands are wiring themselves to volcanoes, and billions keep stacking behind the infrastructure layer. The real question isn't who wins the model war. It's where differentiation moves now that hedging the bottom has become standard.

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide

“Why wouldn’t you want to be in both Pepsi and Coke?” one VC told WIRED. Fair line, but it gives away the game: a lot of investors no longer see foundation models as a winner-take-all bet. They see a category they need exposure to.

Roughly 90 firms and money managers now back both OpenAI and Anthropic. OpenAI shares about 42% of its investor base with Anthropic, and about a third of Anthropic’s backers also hold OpenAI. Some overlap is traditional VC. About 30 are hedge funds, private equity firms, or wealth managers. A few also backed xAI.

For founders, the signal is simple: if capital is hedging at the model layer, you do not get points for picking the “right” lab. Your moat cannot be “we use Anthropic” or “we’re built on OpenAI.” The same investors own both sides.

That pushes the burden of proof up the stack. Real differentiation now means proprietary data, workflow ownership, distribution, or vertical depth. If the base models are becoming a basket, the application layer is where conviction still has to be earned.

🤖 Gusto launches an AI-powered “co-founder” to automate payroll and compliance for over 300,000 businesses.

The pitch isn’t replacing a founder so much as compressing the back-office stack into a chat box, turning messy edge cases like multi-state compliance into API calls instead of late-night Googling. Early users are already poking at how far it can go beyond payroll, from benefits ops to vendor onboarding.

🔥 Six decades, 800,000 workers: lack of manager support tops burnout causes; leaders can spot and fix it fast.

The fix isn’t free pizza; it’s redesigning how managers run 1:1s, feedback, and workload so people feel seen and resourced, which the 60-year meta-study links directly to performance and retention.

🧑‍💼 95% of founders make key leadership mistakes early; managing people is harder than starting up.

The pattern is brutal: overpromoting early hires, dodging hard feedback, and hiring for loyalty over competence shows up in nearly every postmortem, and the same themes run through Inc’s breakdown of “learning the hard way.”

💡 Nvidia bets $6.5B on photonics tech to boost AI speed and efficiency, aiming to outpace energy bottlenecks.

The bigger tell is where that $6.5B lands: Lumentum, Coherent, Marvell, Corning and Ayar Labs span components, fibers and fully integrated photonic chips, turning Nvidia from a chip buyer into a capital allocator for the entire optics stack supporting AI data centers.