The Sentence That Broke the Chess Board

May 6, 2026. San Francisco. Code with Claude. Ami Vora, Anthropic's new Chief Product Officer, is ten minutes into the keynote when she drops the line that empties the air out of the room:

"We're partnering with SpaceX to use all of the capacity of their Colossus data center."

That's Simon Willison's live blog, captured at 09:12. Then she moves on to feature announcements like she hasn't just rewritten the alliance map of the industry.

Four months earlier, on January 12, Anthropic cut off xAI's access to Claude after discovering Musk's lab had been using it through Cursor for internal development. xAI co-founder Tony Wu sent an internal memo that morning, first reported by Kylie Robison:

"According to Cursor, this is a new policy Anthropic is enforcing for all its major competitors."

Three months before that, Musk had publicly called Claude "misanthropic" on X. Op-eds. Lawsuits in adjacent territory. By any honest reading, mutual hostility.

Today they share a power grid. Fortune put a price tag on the reconciliation in a single headline: "Elon Musk called Anthropic 'evil' 3 months ago. Now he's taking $4 billion to become its data landlord." Business Insider put the volume at 300 megawatts at SpaceX's Colossus One.

Press releases don't stage cease-fires this fast unless something underneath is on fire. The fire has a name: compute.

Three Ingredients (Credit Theo, Then Extend)

I didn't have a clean way to think about this until Theo Browne's video on May 6 crystallized it. The framework is his; the angle below is mine.

There are exactly three ingredients to building a frontier AI lab in 2026, and right now nobody has all three at the scale they need:

1. Research — the people who know which paper is worth reading, which architecture is worth chasing, and which apparent breakthrough is just test-set contamination.
2. Data — the high-grade stuff. The "no, you missed step 2 — fix it" correction traces from people building real software with real frustration. You can buy text. You cannot easily buy that.
3. Compute — gigawatts, not GPUs. The unit changed in the last eighteen months and most of us missed it. When SemiAnalysis describes Colossus 2 as "the first gigawatt datacenter in the world," that's the new benchmark.

A quick scoreboard:

Rendering diagram...

Once you see those columns, every move on the board over the last six months snaps into place. The columns also reveal a fourth ingredient money cannot buy: time. You can write a $50 billion check this morning. You cannot write one that conjures eighteen months of substation construction, transformer lead times, and grid-interconnect approvals. That constraint is the silent backdrop to everything below.

Start with the lab feeling the squeeze most.

Anthropic: The Lab That Tripped Over Its Own Demand

In the first quarter of 2026, Anthropic grew 80x. They had planned for 10x.

That's Dario Amodei's framing, on the same Code with Claude stage. Per Business Insider, the CEO said the company had 80x year-over-year growth in revenue and usage in Q1, and "added, half-joking, that he hopes this doesn't continue because that level of hyper-growth is 'too hard to handle.'" Then the line that explains everything that followed: Anthropic "had planned for anywhere from a 'little' revenue growth to 10x… and that gap between expectations and reality is why his company's computing resources have been stretched thin this year."

The numbers are stark. From Anthropic's April 6 announcement:

• Run-rate revenue surpassed $30 billion, up from ~$9 billion at the end of 2025.
• Over 1,000 business customers each spending more than $1 million annualized — doubled from 500 in February. In two months.
• API volume up 17x year-on-year (Vora at Code with Claude).

Here's the part I'm hesitant to write, because I'm a fan and a daily user. I was one of the people getting throttled.

Through March and April, Claude Code on my Pro plan started dropping context mid-session, returning oddly clipped responses, and burning through quotas faster than the same workloads had two weeks earlier. I assumed it was me. It wasn't. Fortune's April 24 story walked through the postmortem: three engineering missteps — a March 4 reduction in default reasoning effort from "high" to "medium," a March 26 bug that made the model discard its own reasoning history mid-session, and an April 16 system prompt that capped responses at 25 words between tool calls.

All three were resolved by April 20. But the underlying story was the one Anthropic conceded to Fortune in plain English:

"Demand for Claude has grown at an unprecedented rate, and our infrastructure has been stretched to meet it, particularly at peak hours."

A number gives the regression teeth. Veracode's analysis found Claude Opus 4.7 introduced a vulnerability in 52% of coding tasks tested, against roughly 30% for OpenAI's models. For a window of weeks, Opus 4.7 wasn't just slower at peak hours — it was measurably worse at the one thing Anthropic's reputation rests on.

And it isn't fixable on a quarter timeline. As MindStudio put it in their April 23 piece: "Ordering GPUs, signing colocation deals, and provisioning infrastructure takes 18 to 24 months at minimum. Money raised today turns into compute capacity in late 2026 or 2027."

Anthropic's compute portfolio reads in that light. Each row is a phone call:

• Google + Broadcom (Apr 6, 2026): multiple gigawatts of next-generation TPU capacity, online from 2027 (Anthropic official).
• CoreWeave (Apr 10, 2026): a multi-year deal, compute online later 2026 with option to expand.
• $50 billion U.S. AI infrastructure commitment (Nov 2025), referenced in Anthropic's announcement.
• AWS Trainium (primary cloud provider and training partner, per Anthropic).
• SpaceX Colossus (May 6, 2026): the new news.

Read those dates: April 6, April 10, May 6. A five-week sprint to bolt anything that produces electrons onto the side of the company. If you've already sold next year's GPU capacity twice over, you make a phone call you swore you'd never make.

xAI: The Lab With Spare Compute and a Data Hole

Memphis has the opposite problem.

Colossus 1 — the original xAI supercomputer, built in 122 days from a converted Electrolux factory — runs at roughly 300 MW with ~200,000 H100/H200s. SemiAnalysis calls it "the largest fully operational, single-coherent cluster" anywhere. It's also mostly idle for Grok; adoption never hit the trajectory Musk projected. xAI's own training has moved next door to Colossus 2, where the January 2026 expansion targets ~555,000 Nvidia Blackwell units and ~2 gigawatts (though Tom's Hardware notes satellite imagery shows only ~350 MW of cooling, so treat the headline as ambition).

The "genius move" SemiAnalysis flags: Memphis pushed back hard on gas-turbine permits, so xAI bought a former Duke Energy plant in Southaven, Mississippi — across the state line, two miles south — where regulators granted temporary approval to run gas turbines for up to 12 months without a permit. Power that would have taken three years of Tennessee paperwork came online in months. Infrastructure as regulatory arbitrage. That's the new game.

So xAI has compute. What it lacks is data — the kind that matters most for the coding-agent market the labs are now fighting over.

Hence April 21, 2026: SpaceX announces a $10 billion collaboration with Cursor to develop "coding and knowledge work AI," plus an option to acquire Cursor outright for $60 billion later in the year. Per TechCrunch, the partnership pairs "Cursor's product and distribution to expert software engineers" with SpaceX's Colossus, "which the company claims has the equivalent compute power of a million Nvidia H100 chips."

Cursor's valuation arc is its own short story:

• January 2025: $2.5 billion
• May 2025: $9 billion
• November 2025: $29.3 billion (post-money on a $2.3 billion Series D)
• April 2026: $50 billion target for the next round
• April 2026: $60 billion option price from SpaceX

One company, 24x in 16 months.

Here's the read I'll commit to as a builder even if I'd hedge it as an analyst: the $10 billion is a data-licensing fee dressed as a partnership. Cursor's value isn't the IDE. The IDE is good; the IDE is not $60 billion good. What's $60 billion good is the corpus of every developer correction message ever sent to a frontier coding model — and Cursor has them against Claude, GPT, and Gemini. Those "no, you missed step 2 — fix it" messages are the highest-grade reinforcement-learning training data on Earth. Each one teaches the next model to skip the step that triggered the correction.

That's why Anthropic banned xAI from Claude in January. Not pettiness. Anti-leakage. Anthropic was the first to admit out loud that competitor IDEs are data faucets, with terms that explicitly prohibit using Claude to "build a competing product or service, including to train competing AI models." Three months later, SpaceX put a price tag on the water. And the plumbing was already in: two months before the deal, two senior Cursor engineers, Andrew Milich and Jason Ginsberg, had left to join xAI, both reporting directly to Musk.

OpenAI's Quiet Wedge: AWS Distribution

OpenAI doesn't need compute. So why did it sign a deal with Amazon?

Until early 2026, Anthropic's enterprise wedge was almost embarrassingly simple: Claude was the only frontier model on Bedrock, and most Fortune 500 procurement runs through AWS. A procurement moat OpenAI couldn't reach. Then in February 2026, AWS launched a multiyear partnership with OpenAI — per CIO Dive — "to distribute OpenAI Frontier, the Anthropic competitor's enterprise platform for AI agents." The signal matters more than the mechanics: the Bedrock moat is leakier than a year ago.

Couple that with the Codex catch-up. The same Fortune piece slipped in a devastating data point: "OpenAI said it now had 4 million active Codex users, 9 million paying business customers, 900 million weekly active users on ChatGPT, and more than 50 million subscribers. Anthropic has not published comparable user figures."

Now you can read why Anthropic broke bread with the man who called them evil. Two of their three moats — coding leadership and Bedrock-only distribution — were narrowing at once, and the only operator with significant idle frontier-grade compute happened to run a competing lab and hate their CEO on principle. When OpenAI doesn't need compute, it buys distribution; when Anthropic doesn't have compute, it leases from its enemy. Each plays the move that costs it the least pride.

Fascinating geopolitics — none of which matters until you sit down at your keyboard.

What This Looks Like at 8 a.m. in Singapore

If you're reading this on a Pro or Max plan in Singapore on Thursday morning, here's what changed for you on Wednesday. From Simon Willison's live blog, captured at 09:12:

• Claude Code's 5-hour session limit doubled for Pro, Max, and Enterprise.
• Peak-hour rate-limit reductions removed.
• "Increased rate limits for developers on Claude Code and the API."

That's not generosity. It's triage — what happens when a 300 MW chunk of Colossus 1 takes traffic off the bottleneck. Real relief. But read against the 80x quote, it's opening a second lane on a road running at 800% over design capacity.

Three things this should change about how you build, Monday morning:

1. Treat compute as a multi-year supply chain, not a budget line. Anthropic's run-rate jump from $9B to $30B doesn't translate into capacity for 18–24 months. The Google/Broadcom TPUs come online "starting in 2027"; the CoreWeave compute "later in 2026." The capital is real; the electrons take time. Implement exponential backoff like you mean it, and treat rate limits as a permanent feature of the landscape, not an inconvenience until your next plan upgrade.

2. Assume your IDE is a training corpus. Every "no, you missed step 2" you type into Cursor, a Claude Code session, Antigravity, any agentic tool — that's training data for somebody. The Cursor deal made the price tag explicit; it didn't invent the dynamic. If your code is sensitive, or your differentiation lives in your prompts and domain knowledge, read the data clauses again. Then read them a second time.

3. Pick your dependencies with the three ingredients in mind. A provider strong on research and data but compute-starved (Anthropic, today) will throttle you when demand spikes — its own statement about being stretched at peak hours is now on the record. One strong on compute but thin on research and data (xAI) will be cheap but lag the frontier. There's no neutral choice, only an informed one. Multi-LLM routing with task-aware fallback isn't a nice-to-have anymore. It's resilience.

Honest admission: I drafted most of this in a Claude Code session, hit a rate limit two thirds of the way through, swapped to a smaller model to finish a section, then came back to Opus to polish. The three ingredients aren't theoretical for me. They're operational. Wednesday's bumps will help. They will not solve it.

Two years ago the bottleneck was can the model do it? One year ago, can humans verify what it produces? In May 2026 it's which lab has enough electricity? — and the answer is that the labs trade favors across grudges most of us would call unforgivable. If you're building on AI, you're not picking a vendor; you're picking an alliance, in a market where alliances reshuffle every quarter and rate limits move faster than your architecture diagrams.