The Anthropic–SpaceX partnership: A new era for creators and entrepreneurs

The AI company behind Claude has partnered with SpaceX to unlock a massive supercomputer and the effects are rolling out to users right now. No jargon, no filler. Just what you need to know.

What actually happened

On Wednesday, May 6, Anthropic announced it had struck a major agreement with SpaceX to use the full computing capacity of Colossus 1, one of the most powerful AI data centers on the planet, located in Memphis, Tennessee. That facility houses over 220,000 NVIDIA GPUs and more than 300 megawatts of capacity. Anthropic is getting access to all of it within the next month.

Advertisement

Get Started

Littlebird is the AI assistant that thinks like you. It knows and understand what you’re working on, so you can stay focused on what matters.

This partnership is surprising because Elon Musk spent months publicly calling Anthropic "misanthropic and evil." But after spending time with their leadership last week, he changed his tune, writing on X that the team "cared a great deal about doing the right thing" and that "no one set off my evil detector."

For context: SpaceX merged with Musk's AI company xAI earlier this year to form SpaceXAI. They've already moved to a newer facility for training their own models, so leasing Colossus 1 to Anthropic made strategic sense for both sides.

Why Anthropic needed this, badly

If you've been using Claude and keep hitting usage limits after just a few tasks, you're not alone. Demand for Claude has exploded, and Anthropic's infrastructure couldn't keep up. The company even admitted that surging demand had caused "inevitable strain" on reliability, especially during peak hours.

The SpaceX deal, combined with a separate multi-billion dollar agreement with Amazon, is Anthropic's big move to fix that. This isn't just adding servers, it's about keeping up with an AI arms race that's accelerating faster than anyone predicted.

What's changing for users right now

These aren't future promises, they're already live for paid subscribers:

• Peak-hour usage limits removed for Claude Pro & Max
• Claude Code rate limits doubled for Pro, Max, Team & Enterprise plans
• API rate limits for Claude Opus models significantly raised for developers

If you pay for Claude, you should notice it getting less frustrating to use, immediately.

Advertisement

Granola — The AI Notepad for back-to-back meetings

Granola helps you keep momentum on what matters. You stay present, Granola turns your conversations into a rich, searchable memory.

GranolaNat Friedman

Claude can now "dream" and that's worth paying attention to

Announced the same day, a new experimental feature called "Dreaming" is now in research preview. Normally, AI has no memory between your sessions, every conversation starts from scratch. With Dreaming, Claude can review previous sessions, spot patterns in how you work, and update its understanding of your preferences automatically.

Think of it as Claude slowly getting to know you. The more you use it, the better it gets at anticipating what you need — without you having to re-explain your business, voice, or workflow every single time.

What this means for creators and entrepreneurs

Content creators: No more being cut off mid-project. You can now run longer writing sessions, batch-generate content ideas, and use AI workflows without constantly hitting walls.

Builders and developers: Doubled rate limits on Claude Code is the headline here. If you're building tools or automations on top of Claude's API, you just got significantly more runway per month.

Solo entrepreneurs: The "Dreaming" feature is especially relevant for you. Over time, Claude will learn your business context, tone, and preferences — making it a far more efficient partner for daily tasks.

Advertisement

Wispr Flow | Effortless Voice Dictation

Flow makes writing quick and clear with seamless voice dictation. It is the fastest, smartest way to type with your voice.

Effortless Voice DictationReid HoffmanCofounder of LinkedIn, Partner at Greylock

International users: A separate Amazon deal in progress will bring new capacity in Europe and Asia, meaning better performance for users outside the US, particularly in regulated industries.

The part most people are ignoring: AI in orbit

Buried in the announcement is something genuinely unprecedented. Anthropic and SpaceX have expressed mutual interest in building multiple gigawatts of AI compute capacity in space. This is a core part of SpaceX's upcoming IPO story, potentially the largest IPO in corporate history. Orbital data centers could eventually offer compute power unconstrained by geography, real estate, or local power grids.

For creators and entrepreneurs, this is a 5–10 year signal: the infrastructure powering your AI tools is about to get radically bigger and more distributed. The ceiling is moving up fast.

One thing worth watching: Musk said SpaceX "reserves the right to reclaim the compute" if Claude's AI ever "harms humanity." Easy to dismiss, but it's a reminder that every AI partnership right now is happening inside a volatile political and competitive environment. Anthropic is also currently in a legal battle with the Pentagon after being blacklisted as a "supply chain risk." The deal is a win. The landscape around it is still complicated.

Actionable Takeaways

For creators:

• If you’re on Pro or Max, start stress-testing longer, more ambitious workflows this week.
• Experiment with multi-agent setups you previously avoided.
• Document what becomes possible now that wasn’t before — those learnings will compound.

For entrepreneurs:

• Audit your current AI usage limits and identify projects you deprioritized because of them.
• Revisit your product roadmap: What could you build in the next 90 days with 2x the Claude Code capacity?
• Consider how orbital-scale compute (even conceptually) changes your 3–5 year vision.

For everyone:

• The era of “AI is capped and expensive to use heavily” is ending faster than expected. The winners will be those who adapt their thinking and operations to abundance.

This partnership is a reminder that in AI, strange bedfellows often make the most powerful alliances and the real breakthroughs come when infrastructure scales dramatically.

The compute walls are coming down. The question is: What will you build now that they’re lower?

I’d love to hear from you:

• What new workflows are you most excited to try with the increased limits?
• Are you rethinking any product or content ideas because of this?

Reply and let’s discuss.