04-26-2026Price:

The Frontier

Your signal. Your price.

AI & TECH

Elon buys Cursor to win AI coding war

Sunday, April 26, 2026 · from 4 podcasts
  • SpaceX secures $60B option to buy Cursor, pairing its AI models with a top coding platform.
  • The deal gives Musk access to developer data needed to train self-improving AI.
  • XAI lags behind rivals, and Musk is betting on tools, not just compute, to catch up.

SpaceX’s move to acquire Cursor for $60 billion isn’t about building better software. It’s about building the machine that builds the software. The deal gives Elon Musk direct access to the logs and workflows of thousands of developers - real-world data that’s essential for training AI to code like a human. Without it, xAI’s models remain stuck in simulation.

Six weeks after the layoffs at Meta and Microsoft, the shift is no longer theoretical. Junior developers are already being displaced by AI that writes, debugs, and deploys code faster than any entry-level engineer. At Anthropic and OpenAI, models now generate the majority of new code. The career ladder is gone. There’s no apprenticeship when the AI does the grunt work.

Cursor’s Composer 2 model already outperforms most commercial coding tools. By integrating it with SpaceX’s Colossus GPU cluster, Musk gains a closed loop: AI writes code, runs it on massive hardware, learns from the results, and rewrites itself. That’s the recursive improvement xAI has been missing. OpenAI and Anthropic are bottlenecked by compute. Musk isn’t.

"The goal is to create a recursive loop where AI researchers use the best tools to build even better AI."

- Alex, Moonshots with Peter Diamandis

The acquisition also solves xAI’s biggest problem: mindshare. Despite massive compute, Grok hasn’t gained traction among developers. Cursor has the user base. It has the interface. Now, it has the backing to go all-in on agentic development. The $10 billion partnership fee SpaceX is paying isn’t a cost - it’s insurance against falling further behind.

This isn’t just about coding. It’s about control. If AI can design, test, and deploy its own updates, the humans are out. Nathaniel Whittemore put it plainly: GPT Images 2.0 already generates functional UIs with working barcodes. Codex turns those into code. The stack is complete. The only missing piece was the data pipeline - now Musk has it.

"If the models are as dangerous as the company suggests, they shouldn't be accessible via educated guesses at a vendor."

- Nathaniel Whittemore, The AI Daily Brief

The era of human-led software development is ending. Musk isn’t waiting for it to happen. He’s buying the tools to make it happen on his terms.

Source Intelligence

- Deep dive into what was said in the episodes

4/23/26: Navy Sec Fired, WH Freaks Over Intel On Iran Military, Food Inflation SpikesApr 23

Also from this episode: (16)

Other (16)

  • Defense Secretary Pete Hegseth fired Navy Secretary John Phalin amid a massive naval blockade of Iranian ports, part of a month-long campaign to remove some 34 top officials Hegseth perceived as disloyal.
  • Axios reported John Phalin's firing stemmed from his refusal to follow orders and conflicts with Pete Hegseth, who saw Phalin's direct access to President Trump as a threat to his authority.
  • The administration named Hung Cao, a former Senate candidate who lost to Tim Kaine and campaigned on fighting 'witchcraft' in Virginia, as John Phalin's replacement for Navy Secretary.
  • Saagar and Krystal criticized President Trump for inventing 'fake concessions' from Iran, such as claiming he secured the release of eight women, despite human rights organizations reporting no imminent executions and some women already released a month prior.
  • Drop Side reported Iran would only resume talks if Trump extended the ceasefire and ended the naval blockade, with Pakistani mediators expecting the blockade's lift.
  • The Defense Intelligence Agency (DIA) assessed Iran retains thousands of missiles and UAVs, contradicting White House claims that the US military decimated Iran's capabilities in 38 days and annihilated its navy.
  • CSIS estimates indicate the Iran War expended significant portions of US munitions, including 50% of FAD interceptors, 50% of Patriot interceptors, and 30% of Tomahawk missiles.
  • Saagar argued the 38-day Iran War set the US military back at least five years, cost hundreds of billions, and depleted critical munitions, making the US vulnerable in future conflicts.
  • CBS News reported that approximately half of Iran's ballistic missile stockpile, 60% of its Revolutionary Guard naval arm, and two-thirds of its air force remain operational despite US and Israeli campaigns.
  • The Intercept reported the Pentagon erased around 15 wounded US troops from the Iran War casualty list, which Krystal believes reflects the American public's low tolerance for military pain and signals a US loss in the conflict.
  • President Trump ordered the US Navy to 'shoot and kill' small Iranian boats and triple mine-sweeping operations in the Strait of Hormuz, likely in response to a Pentagon assessment that clearing the strait could take up to six months.
  • Saagar projected that a six-month closure of the Strait of Hormuz could cause oil prices to hit $250-300, gas prices to reach $6 per gallon, and lead to blackouts and rationing in Asia.
  • Tracy Alloway of Odd Lots reported food company costs jumped almost 8% year-over-year in March, up from 4.2% in February, driven by higher fuel prices, with further increases expected from fertilizer and plastics.
  • United Airlines' CEO stated fares may need to rise 20% to offset surging jet fuel costs, while Lufthansa cut 20,000 flights, saving 40,000 metric tons of fuel, as Europe faces critical jet fuel shortages.
  • The Trump administration is reportedly nearing a deal to loan Spirit Airlines $500 million and take a significant stake, as the carrier struggles with surging jet fuel prices, exacerbated by its blocked merger with JetBlue.
  • Aluminum faces a 'black swan supply shock' due to the Iran War, threatening industries like transport and construction, as the Middle East accounts for roughly nine percent of the estimated global supply, or seven million metric tons annually.

Iran's AI Supply Chain Threat, Claude vs. SaaS, and Elon's $60B Cursor Bet | EP #249Apr 23

  • Peter Diamandis and Alex predict AI will disrupt industries like legal research (Lexus Nexus), business intelligence (Tableau), medical documentation (Epic), financial modeling (Bloomberg Terminal), and HR/recruiting (Workday).
  • Salim mentions a survey from Anthropic employees, predicting entry-level software engineers and researchers will be replaced by AI within three months, calling coders the "canary in the coal mine" for job displacement.
  • Alex argues the current recursive self-improvement in AI means a wide "blast radius" of displacement, noting rumors of Google DeepMind code being generated by Claude.
  • Elon Musk projects XAI's Grok 4.4 and Grok 4.5 will reach 1 trillion and 1.5 trillion parameters, respectively, predicting Grok 5 to be AGI, followed by ASI and ASI 2 for Grok 6 and 7.
  • Alex speculates that OpenAI's executive departures signify a renewed focus on recursive self-improvement and code generation, potentially leading to the emergence of a new frontier AI lab.
  • Peter Diamandis announces the release of ChatGPT Images 2.0, an image generation model with 99% text accuracy, extraordinary resolution, and web search capabilities for generating infographics and solving math problems.
  • Alex notes ChatGPT Images 2.0 achieved an ELO score of 1,512 in arena.aI benchmarks, making it the top-scoring text-to-image model. He speculates its release might indicate lower compute costs or its instrumental role in code generation.
  • SpaceX is negotiating the right to acquire Cursor for $60 billion, which Dave suggests will provide XAI with needed foundation models, compute power, and distribution for code generation capabilities.
  • Alex views SpaceX's potential acquisition of Cursor as a sign that code generation is the core focus for frontier AI labs, with Elon possibly using Cursor's expertise to reboot Grok's code generation capabilities.
  • Alex argues that the data center buildout, currently private-funded, will become the de facto economy, consuming a significant portion of GDP, as it is foundational infrastructure for future civilization and AI intelligence.
  • Alex predicts that future technological advancements, including advanced nanotechnology, will enable sovereign nation-states to domesticate their entire supply chains, diminishing the long-term future of global supply chains.
  • Peter Diamandis reports that Blue Origin's New Glenn rocket achieved its third successful booster reuse, landing beautifully. However, its AST satellite was launched into the wrong orbit and needs de-orbiting.
Also from this episode: (11)

Science (1)

  • Peter Diamandis states that the annual Breakthrough Prize, co-founded by Yuri Milner, Sergey Brin, and Mark Zuckerberg, was established in 2012 and awards $3 million for scientific advancements.

Models (4)

  • Peter Diamandis reports that Anthropic's Claude Design release led to Figma's stock dropping 10% and Adobe's dropping 2%. Alex considers this an "unhobbling" of existing latent model capabilities.
  • Dave emphasizes Anthropic's need for better user experience and an independent software vendor (ISV) strategy, criticizing its current lack of roadmap and speed, despite powerful capabilities.
  • Alex critiques Elon Musk's focus on raw parameter counts, arguing that the industry should prioritize "intelligence density" by compressing more capability into smaller models, not just scaling up.
  • XAI launched a standalone speech API with a 5% error rate on phone calls, outperforming 11 Labs (12% error rate). Peter Diamandis highlights its competitive price of 10 cents per hour and support for 25 languages.

Psychology (1)

  • Alex argues that the idea of human attention as a scarce resource is incorrect; he believes humans can 'manufacture' more attention. Peter Diamandis expresses interest in this concept.

Enterprise (1)

  • Salim believes big companies will struggle to adopt AI natively due to human-centric workflows, ultimately acquiring AI-native startups to compete and redesign their operations.

AI Infrastructure (4)

  • Peter Diamandis notes XAI's compute power leadership, with an estimated 2 gigawatts by year-end, surpassing OpenAI (1.2 gigawatts) and Meta/Anthropic/AWS (1 gigawatt each).
  • Salim suggests that tech companies building AI applications should ensure their tech stack is model-agnostic to easily swap underlying AI models and maintain competitive edge through customer relationships and niche expertise.
  • Alex explains that SpaceX is transforming into a hyperscaler, building a "Dyson swarm" of orbital clouds. The Cursor acquisition could signify an opening of SpaceX's vast ground and orbital GPU sets to third parties.
  • Peter Diamandis notes that data center CAPEX is projected to hit nearly $1 trillion over six years, dwarfing historical government-funded megaprojects like the Apollo program ($257 billion over 14 years) and Manhattan Project ($36 billion over 5 years).

SpaceX and Cursor team up to topple Claude Code | E2279Apr 22

  • SpaceX bets billions on Cursor to secure the data needed for recursive AI self-improvement.
  • Bitstarter launches a crowdfunding model to strip power from predatory Bit Tensor investors.
  • Subnet 11 creates a sandbox for AI agents to write their own instruction sets.

What GPT Images 2 UnlocksApr 22

  • SpaceX partnered with Cursor, an AI coding tool, acquiring rights to purchase Cursor for $60 billion later this year; if the acquisition fails, SpaceX will pay Cursor $10 billion for their collaborative work.
  • XAI could benefit from Cursor by gaining a significant data pipeline to improve its models, especially since XAI has struggled to generate revenue or release impactful models, and lacks a footprint in the AI coding space.
  • Sam Altman criticized Anthropic's promotion of Mythos, suggesting its fear-based marketing positions AI control as a justifiable purchase, rather than focusing on legitimate safety concerns.
  • Google released an upgrade to its Deep Research agents, now featuring MCP support for third-party data and the ability to output charts and infographics using Nano Banana models, with a Max version outperforming GPT 5.4 and Opus 4.6.
  • The improvements in Google's Deep Research agents, despite still using Gemini 3.1 Pro under the hood, stem entirely from harness upgrades and additional inference, not a more advanced base model.
  • GPT Images 2.0 offers enhanced precision and control, handling small text, UI elements, and dense compositions at resolutions up to 2K, along with multilingual capabilities for designs where language is integrated.
  • Nathaniel Whittemore argues ChatGPT Images 2.0 is the first image model for the 'agentic era' because its primary impact will come from integration with other systems, rather than standalone viral moments.
  • Users are already integrating GPT Images 2.0 with Codex, creating a pipeline to generate UI mockups and then convert them into working code, addressing Codex's previous limitations in UI design.
Also from this episode: (5)

Startups (1)

  • The SpaceX-Cursor deal potentially solves Cursor's reported issue of losing money on every Claude and OpenAI token served, giving them access to XAI's Colossus training supercomputer with millions of H100 equivalent units for in-house model development.

Markets (1)

  • SpaceX's IPO disclosure documents reveal Elon Musk increased his stake by $1.4 billion and could receive a compensation package tied to market cap achievements ranging from $1.1 trillion to $6.6 trillion.

Safety (1)

  • An unauthorized group accessed Anthropic's Claude Mythos preview via a third-party vendor and information from the Merkle data breach, despite Anthropic's tight control measures for cybersecurity purposes.

Models (2)

  • OpenAI's new ChatGPT Images 2.0 model leads the Arena Elo score human preference board with a record-breaking 242-point lead over the previous leader, indicating a significant jump in quality.
  • While GPT Images 2.0 shows vast improvements, Boyan Tongues noted visual artifacts, and Sharon Goldman's sister found anatomical inaccuracies in medical images, highlighting a zero-tolerance for errors in certain use cases.