← All episodes

SpaceX and Cursor team up to topple Claude Code | E2279

| 8 products mentioned
Watch on YouTube ai coding assistants startup funding decentralized networks bittensor subnets compute infrastructure ai model competition venture capital

SpaceX's partnership with Cursor to build advanced AI coding models represents a strategic consolidation play that could reshape the entire AI development landscape—combining Cursor's market-leading Composer model and developer data with XAI's computational firepower at a potential $60 billion acquisition price by end of 2026. The episode also explores how BitTensor subnets are creating decentralized markets for AI innovation, with new platforms like Bitstarter democratizing access to subnet launches and Trajectory RL building competitions to improve AI agent skills, suggesting that the future of AI development may be as much about distributed networks as centralized companies.

Key takeaways
  • The SpaceX-Cursor deal effectively gives XAI access to production-level developer data and proven model-building expertise while signaling confidence ahead of SpaceX's IPO, positioning the company to compete with Claude Code and CodeEx if execution succeeds.
  • BitTensor subnets solve information asymmetry problems in decentralized AI launches by offering retail investors the same OTC-style terms as VCs, with Bitstarter reducing fundraising requirements from $250K+ down to whatever validation a team can secure through crowdfunding.
  • The economics of subnet creation heavily favor early movers—a subnet slot costs 0.25M+ TAO, subnet owners only capture 18% of emissions (41% each to miners and validators), and most founders need VC backing that extracts 20-30% in perpetuity, creating a capture problem Bitstarter aims to solve by taking only 3% for 90 days.
  • Skills (markdown-formatted instruction files for AI agents) represent an emerging product category that can be competitively improved through decentralized networks like Trajectory RL, which runs monthly seasons where agents compete to write better skills, already showing measurable improvements in self-learning skills within one week of launch.
  • Compute constraints are forcing every major tech company to build proprietary AI coding tools—Google's new TPU8T (training) and TPU8I (inference) chips show Nvidia won't monopolize chip supply, while Anthropic's $100B/10-year AWS commitment signals the race for exclusive capacity is accelerating faster than model improvements.
  • Trajectory RL's vision of AI agents writing skills instead of humans opens a path to recursive improvement (agents improving themselves) and recurring monetization through data sales and fine-tuning, potentially creating a higher ceiling than hand-written skills as competitions drive more sophisticated solutions.

Recommendations (4)

BitTensor
BitTensor uses

"Bittensor is adversarial and the miners are like very very intense. They're going to tear you apart."

Chris Zakaria · ▶ 22:20

OpenClaw
OpenClaw uses

"Open Clock Mania is dying down. That's how I feel. People might want to take a second look at Open Clock."

This Week in Startups · ▶ 1:07

Claude
Claude uses

"Opus 4.7 tooled up to the max code and co-work 247. I'm like consistently on like three instances of Claude code and composer and like making them fight against each other."

Chris Zakaria · ▶ 17:35

MacBook
MacBook uses

"My pink MacBook Neo does not have special anti-soda properties to it. I paid like 600 bucks for it, which for a laptop."

This Week in Startups · ▶ 1:11:11

Mentioned (4)

Cursor
Cursor "SpaceX and Cursor are partnering on AI models. Of course, Cursor, the popular AI coding model and..." ▶ 3:39
Composer "They released Composer 2, which is their latest model that they built for themselves. It was anno..." ▶ 4:59
Codex
Codex "Cursor has done a very good job competing with Codex from OpenAI and also Claude Code from Anthro..." ▶ 4:40
Claude Code
Claude Code "Cursor has done a very good job competing with Codex from OpenAI and also Claude Code from Anthro..." ▶ 4:43