Five Stories That Actually Mattered This Week
The signal behind the noise. No hype, no fluff.
There's a lot of noise in startup news most weeks. This week had some actual signal. SpaceX filed an S-1 that reads like science fiction with real financials. Anthropic is about to surpass OpenAI as the most valuable private company in the world. And the venture market is quietly telling founders something important about what gets funded in 2026. Here's what matters.
1. SpaceX Filed Its S-1. The Numbers Are Extraordinary and the Risks Are Real.
SpaceX filed its S-1 on May 20, targeting a valuation of approximately $1.75 trillion. The company generated roughly $18 billion in revenue in 2025 and is seeking what would be the largest IPO in American history. Elon Musk's compensation package is tied not just to stock price milestones but to the creation of a permanent human colony on Mars with at least one million inhabitants. His existing stake alone, at the expected IPO valuation, would be worth roughly $700 billion.
Spence's Take
A $28 trillion TAM pitch and a Mars colony compensation trigger in the same S-1 is the most audacious investor document I've ever read. The Starlink numbers are real. Everything else is a very expensive hypothesis.
Mental model: Barbell Strategy.
SpaceX's financial reality is itself a barbell. Starlink's satellite internet division posted an operating profit of $4.4 billion, making it the only segment generating consistent cash. That's the safe 90%. The other side of the barbell — Starship, Mars, orbital data centers — is the wildest long-duration bet in corporate history. Investors who understand this aren't buying the Mars story. They're buying a profitable satellite internet business with an optionality package attached. The risk is that the market prices the Mars story and the Starlink reality gets ignored in the valuation math. At 94x 2025 revenue, there's not much margin for the hypothesis to be wrong.
2. Anthropic Is About to Become the Most Valuable Private Company in the World.
Anthropic is set to close a round that may top $30 billion at a valuation above $900 billion as soon as next week, co-led by Sequoia, Dragoneer, Altimeter, and Greenoaks. That would vault it ahead of OpenAI, currently valued at $852 billion. The company reported $4.8 billion in Q1 revenue and is projected to hit $10.9 billion in Q2, with an annualized run rate expected to exceed $50 billion by the end of June.
Spence's Take
Q1 to Q2 revenue more than doubling in a single quarter, while already at $5 billion quarterly, is one of the most extraordinary revenue curves I've seen. The valuation isn't the story. The trajectory is.
3. The SpaceX S-1 Revealed Anthropic Is Paying $15 Billion a Year for Compute.
Buried in the SpaceX filing: Anthropic is paying SpaceX approximately $15 billion annually. This is a direct commercial relationship between two companies whose founders have a famously adversarial public history. Musk once called Anthropic "evil." That's now apparently a $15 billion customer relationship.
Spence's Take
When ideology and economics go head-to-head, economics wins. Every time. The SpaceX-Anthropic commercial relationship is the cleanest possible proof of that principle.
4. Hark Raised $700 Million Before Shipping a Single Product.
Brett Adcock's AI lab Hark closed a $700 million oversubscribed Series A at a $6 billion post-money valuation this week, led by Parkway Venture Capital with participation from NVIDIA, AMD Ventures, Intel Capital, Qualcomm Ventures, Salesforce Ventures, ARK Invest, and Brookfield. Adcock self-funded the company with $100 million of his own money before raising a dollar from anyone else. The company has roughly 70 employees, no shipped product, and plans to release its first AI models this summer.
Spence's Take
Founder-credibility-based fundraising at this scale requires the first product to validate the thesis publicly. The summer software release will tell you everything about whether the $6 billion is justified.
5. The VC Market Is Concentrating at the Top. Fast.
US startups have raised $303 billion in 2026 to date, a 130% increase on the same period in 2025, but the distribution is increasingly uneven. The top five rounds this year account for a disproportionate share of total capital deployed, and seed and early-stage deal counts continue to decline even as total dollars rise.
Spence's Take
More capital, fewer deals, sharper selection. If your Series A story doesn't stand out in the first two minutes of a meeting in this market, you're not getting to minute three. The bar isn't higher because investors are nervous. It's higher because they have more money and less time.