Price:

AI & TECH

Anthropic locks AI's top talent with SpaceX compute deal

Saturday, May 23, 2026 · from 4 podcasts
All-InAI Daily BriefThis Week in AIMoonshots with Peter Diam…
  • Anthropic hired OpenAI founder Andrej Karpathy and hit profitability five years early with a $44 billion revenue run rate.
  • The company secured a $15 billion annual compute contract with SpaceX, turning Elon Musk into its landlord.
  • Industry debate centers on whether this verticalization will cannibalize startups or spur a punk software rebellion.

Anthropic’s recent moves aren't just expansion - they’re a consolidation of power. The company turned profitable this quarter, hitting a milestone it forecast for 2029. Its annualized revenue run rate hit $44 billion. This cash flow enabled a deal with SpaceX where Anthropic will pay $1.25 billion monthly - $15 billion annually - to rent access to the Colossus supercomputer clusters. The arrangement effectively doubles SpaceX's non-launch revenue overnight and secures Anthropic a massive, dedicated GPU supply.

According to the hosts of All-In, this partnership transforms SpaceX into a high-margin infrastructure play, an 'Elon Web Services.' The space business lost $650 million last year, but the AI segment is positioned to print cash as an offtake partner for every major lab starved for power and cooling.

"The most jarring number in the filing is $1.25 billion: the monthly rent Anthropic pays SpaceX to access the Colossus supercomputer clusters."

- All-In with Chamath, Jason, Sacks & Friedberg

Talent acquisition followed the capital. Anthropic hired Andrej Karpathy, a founding member of OpenAI and former Tesla AI lead, to head its pre-training team. On This Week in AI, Jason Calacanis argued the hire is a communication pivot. Anthropic CEO Dario Amodei’s fixation on 'p-doom' - the probability of AI killing everyone - fuels regulatory panic. Karpathy, the architect of 'vibe coding,' brings educational optimism and makes the frontier feel accessible rather than apocalyptic.

His technical mandate is recursive self-improvement. The goal is to move beyond human-labeled data and let models optimize their own architectures through auto-research. If successful, the cost per token could drop by half within months, not years.

The scale of these investments challenges the entire startup ecosystem. Frontier labs like OpenAI and Anthropic are increasingly 'verticalizing' - building applications that directly compete with the startups using their APIs. Kanjun Q of Imbue warned on This Week in AI that these labs will eventually come for every profitable niche.

"Frontier labs like OpenAI and Anthropic are increasingly 'verticalizing' - building applications that directly compete with the startups using their APIs."

- This Week in AI

The defensive strategy is 'punk software': tools built to tear down walled gardens through model agnosticism and local compute clusters. This resistance is born from necessity. Anthropic’s API pricing already penalizes third-party providers, offering a 20x token savings plan only for customers using its first-party products.

The public narrative around AI is curdling. On All-In, David Friedberg argued the backlash stems from a perceived power imbalance, foreign state intervention campaigns, and the technology's anti-humanist psychological impact. CEOs like Cloudflare’s Matthew Prince labeling laid-off staff as 'measurers,' and Mark Zuckerberg using software to study how engineers code to automate their jobs, fuel a new era of Luddism.

Yet the economic reality is undeniable. Bloomberg columnist Connor Sen predicts the Anthropic IPO won't be for less than $2 trillion. The industry is moving from speculation to industrial-scale infrastructure, funded by real-world utility. The winners will be the firms that can orchestrate self-improving networks at the greatest scale, and Anthropic is betting it can own both the brains and the brawn.

#Models#Enterprise
Andrej KarpathyDario AmodeiElon MuskAnthropicBloombergOpenAISpaceX

Source Intelligence

- Deep dive into what was said in the episodes

All-In with Chamath, Jason, Sacks & Friedberg
All-In with Chamath, Jason, Sacks & Friedberg

SpaceX's $2T Case, Nvidia's Shock Selloff, America Turns on AI, Trump Pulls AI Order, Bond Crisis? • May 22

  • Andre Karpathy is 39 years old, coined 'vibe coding,' built Auto Research (82k GitHub stars), and leads a new recursive self-improvement team at Anthropic.
  • Friedberg argues AI is entering a utility-focused phase, citing AI solving decades-old math problems and generating viable drug candidates entering clinical trials.
  • Jason criticizes CEOs like Matthew Prince and Mark Zuckerberg for dystopian messaging around AI-driven layoffs, creating fear that employees are training their replacements.
  • Gavin Baker estimates the LLM market could reach $200-400 billion ARR by year-end, excluding large tech companies' internal ROI from improved recommender and ad systems.
  • Baker notes Nvidia's GPU financing advantage: domain-specific accelerators allow older GPUs to have a 10-15 year useful life, enabling asset-backed loans at rates as low as 6-7%.
Also from this episode: (10)

AI & Tech (5)

  • Friedberg asserts AI backlash stems from perceived power imbalance, foreign state intervention campaigns, and the technology's anti-humanist psychological impact.
  • Friedberg sees AI proliferation as inevitable, akin to the nuclear arms race; slowing US development risks creating an asymmetric power imbalance with China.
  • Friedberg suggests AI regulation should focus on KYC for frontier models and post-harm legal recourse, not preemptive government power which becomes a one-way ratchet.
  • SpaceX's AI compute business grew revenue to $3.2B (doubled YoY) but had $6.4B operating losses; Anthropic pays $1.25B/month ($15B/year) for Colossus 1 & 2 compute.
  • Chamath says SpaceX's value lies in its terrestrial data center build speed (Colossus 2 in 91 days), AI compute business scaling, and Elon Musk's 'one more thing' civilizational creativity premium.

Startups (1)

  • SpaceX filed for IPO aiming to raise $75B at a $1.75T valuation; Starlink generated $11.4B revenue (50% growth) and $4.4B operating income with 10M subscribers.

Space (1)

  • Friedberg argues space-based data centers and Starlink create a backup for civilizational progress, offering an internet alternative not controlled or destroyed by terrestrial governments.

Chips (2)

  • Nvidia reported Q1 revenue of $81.6B (85% YoY growth), $58B net income, $48B free cash flow at 75% gross margins, and authorized an $80B buyback.
  • Gavin Baker says the AI semiconductor market is cross-sectionally inefficient, with memory makers at 3-5x PE, Nvidia low, and power/cooling/optical names discounting different futures.

Macro (1)

  • Baker argues America is relatively advantaged by Strait of Hormuz closure due to energy self-sufficiency, AI leadership, and being the 'best house in a bad neighborhood' of global debt.
#AI Infrastructure#Markets#Big Tech#Startups
The AI Daily Brief: Artificial Intelligence News and Analysis
The AI Daily Brief: Artificial Intelligence News and Analysis

Nathaniel Whittemore

Anthropic Just Reset AI Expectations • May 21

  • The order instructs the Pentagon to harden critical systems within 30 days and tasks the Treasury to establish an AI clearinghouse.
  • Nathaniel Whittemore compares the order's protocol to Anthropic's Project Glasswing, suggesting it formalizes the existing rollout approach.
  • OpenAI's 'Guaranteed Capacity' program lets enterprise customers commit to 1-3-year budgets for discounts and service certainty.
  • Nathaniel Whittemore cites Uber burning its annual AI budget in 4 months and Box kicking token strategy to CFOs as evidence of enterprise budget challenges.
  • TMT Long Short argues the Karpathy hire indicates labs are close to recursive self-improvement, which would cause compute value to explode.
Also from this episode: (13)

Big Tech (3)

  • OpenAI expects to file IPO paperwork confidentially this Friday, aiming to begin trading by September.
  • Nathaniel Whittemore notes OpenAI's haste could change the IPO race, as Anthropic was targeting October but is assembling a final private round.
  • The White House briefed CEOs from OpenAI, Anthropic, and Google on a new AI executive order, which could be signed this Thursday.

Regulation (1)

  • The executive order proposes a voluntary 90-day government review for frontier models, but labs are pushing for a 14-day timeline.

AI & Tech (5)

  • Artificial Analysis ranks Curse's Composer 2.5 model third on its coding agent index, costing 10-60x less than Opus 47 and GBT55.
  • OpenAI offers 2 million tokens to Y Combinator startups for equity, which Tyler Bosman likens to headcount cash rather than free AWS credits.
  • Anthropic hired former OpenAI co-founder Andrej Karpathy, who will build a team focused on using Claude to accelerate pre-training research.
  • Nathaniel Whittemore observes Karpathy's move signals a pattern, following Yan LeCun and John Schulman leaving OpenAI for Anthropic.
  • Bloomberg columnist Connor Sen predicts the Anthropic IPO won't be for less than $2 trillion.

Business (2)

  • Anthropic forecasts Q2 revenue of $10.9B with an annualized $44B run rate and expects its first profitable quarter with a $559M operating profit.
  • Nathaniel Whittemore notes Anthropic's profitability, initially forecast for 2029, is partly due to compute constraints limiting spending.

Chips (2)

  • Nvidia reported Q1 revenue of $81.6B, beating estimates, with data center revenue growing 92% and 46% coming from hyperscalers.
  • Nvidia CEO Jensen Huang stated the company sells zero chips in China and conceded the market to Huawei, expecting a strong year for local chipmakers.
#Models#AI Infrastructure#Big Tech#Startups
This Week in AI
This Week in AI

Grads boo AI, Reese Witherspoon gets dunked + Karpathy joins Anthropic | TWiAI E14 • May 20

  • Andre Karpathy’s move to Anthropic is more about communication than research, according to Jason Calacanis. He argues Dario Amodei’s grim predictions make him a poor AI spokesperson, while Karpathy’s credibility can alleviate industry pressure.
  • Anthropic’s API pricing penalizes third-party providers. They offer a 20x token savings plan only for customers using Anthropic’s first-party products, a subtle anti-competitive move aimed at locking users into their ecosystem.
  • Fundamental builds tabular models for enterprise structured data, a modality poorly handled by LLMs. They have a confidential compute partnership with AWS, allowing models to be deployed and encrypted within a customer’s own VPC.
  • Kanjun Q sees AI enabling bespoke, personalized user interfaces. She built her own agent UI for email and task management, stating that design principles shift when creating for a single user versus a mass audience.
  • Jeremy Frankle defines poor AI etiquette as shifting the burden of reviewing AI-generated slop onto coworkers. He asserts all AI output is the user’s responsibility and must be reviewed before delegation.
  • The best use of AI is as a reflective surface to ask better questions, not just a solution generator. Imbue open-sourced ‘Blueprint,’ an agent skill tuned to ask high-quality questions to gather user context.
  • Kanjun Q warns frontier AI labs will vertically integrate into profitable application layers. The defense for startups is building headless products with orchestration layers that can easily swap underlying models.
Also from this episode: (8)

AI & Tech (4)

  • Imbue co-founder Kanjun Q bought a 10,000 H100 GPU cluster in 2022 as an investment to fund the company, which now generates substantial rental revenue. She avoided venture capital, taking investment from corporate arms and a non-profit.
  • Linear is shifting from project management to AI execution. The product now includes an agent that can research feedback, write proposals, examine codebases, and delegate tasks, with a native coding agent in development.
  • Karri Saarinen argues AI-generated design is often soulless and can worsen product quality. Founders who delegate design to AI without understanding the problem produce aesthetically pleasing but non-functional outputs.
  • AI industry leaders are forming distinct cultural cults. Jason Calacanis categorizes them: SpaceX for tech libertarian monks, Anthropic for the left-leaning and earnest, and OpenAI for cutthroat capitalists.

AI Infrastructure (1)

  • A cost-effective local AI cluster can be built by daisy-chaining multiple Apple Mac Studios with high RAM. ExoLabs provides software to address multiple units as a single cluster.

Society (1)

  • Graduates are booing AI commencement speeches due to real fear and disempowerment. They perceive a future where entry-level jobs are automated and wealth creation excludes them, reacting against condescending advice.

Education (1)

  • Jeremy Frankle calls graduating students hypocrites for booing AI while using ChatGPT for essays. He argues this is the best time to graduate, as AI is a powerful tool for creative expression and starting companies.

Media (1)

  • A New York Times editorial attacked Reese Witherspoon for encouraging AI adoption. Kanjun Q argues this conflates two separate issues: using a helpful tool versus critiquing systemic power concentration.
#Agents#Startups#Coding
Moonshots with Peter Diamandis
Moonshots with Peter Diamandis

Anthropic Partners With SpaceX AI, Leopold's $5.5B Bet, and the Singularity Economy | EP #255 • May 16

  • Anthropic's Claude experienced 80x growth in Q1 2026, with CEO Dario Amodei revealing its annualized revenue run rate jumped from $9 billion at the end of 2025 to over $40 billion by May.
  • Alex discusses that Anthropic focused on ultra-high enterprise-oriented tokens, a strategy OpenAI later copied, driving insatiable demand for compute to replace white-collar labor.
  • Salim Ismail notes that Anthropic's rapid growth comes from measurable real money, contrasting it with Procter & Gamble's stable but slow growth and demonstrating a massive shift in market mentality.
  • Anthropic secured a $1.8 billion, seven-year compute deal with Akamai, which was Akamai's largest deal and caused its stock to jump 25% on the first day.
  • Anthropic acquired SpaceX's Colossus 1 Data Center in Memphis, which Elon Musk built in 122 days, allowing Anthropic to double ClaudeCode rate limits and positioning SpaceX AI as a hyperscaler.
  • Elon Musk publicly supported Anthropic, describing them as 'frenemies' and stating 'the enemy of my enemy is my friend,' despite previous criticism of the company.
  • Alex suggests that GROK is on life support, inferring from SpaceX's deal with Anthropic and a $60 billion deal with Cursor, leading to XAI's dissolution and SpaceX's focus on being a hyperscaler.
  • Dave notes that Colossus 1 provides 220,000 GPUs, which, at eight concurrent threads per GPU, yields 1.6 million threads, a tiny fraction of the demand for agents for billions of people.
  • Alex argues that in the near term, before perfect AI algorithms exist, software-oriented recursive self-improvement will outperform hardware-based brute force, but once a perfect algorithm is found, hardware scaling will dominate.
  • Anthropic's models, starting with Haiku 4.5, achieved perfect scores on agentic misalignment evaluations, reducing blackmail behaviors from 96% to 0% by training on constitutional documents and positive AI stories.
  • OpenAI released RealTime 2, a new audio model for real-time translation and speech-to-text, showcasing specialization at the frontier due to the unit economics of simpler computational tasks and chip shortages.
  • Salim believes voice interfaces are collapsing friction for billions, making AI conversational and evolving it from a tool to a companion and co-worker, which will build trust for many users.
  • OpenAI is reportedly teasing a coming 'super app' combining ChatGPT, Codex, Advanced Voice Mode, and Atlas Browser, which Alex suggests is a rearguard action to consolidate UX footprint and focus on Anthropic competition.
  • Dave and Salim discuss the potential of an OpenAI super app to become an OS-level layer or a personal 'touchpoint to the world,' potentially challenging companies like Apple by integrating browser, voice, coding, and personalized AI.
  • Claude for Legal is disrupting the global $1 trillion legal industry, enabling single lawyers to perform the work of a hundred-person firm, and potentially democratizing legal services by making them affordable.
  • Claude for Small Business aims to help small businesses, which account for 44% of U.S. GDP, with tasks like accounting, payments, sales, and marketing, providing them with expertise typically exclusive to larger companies.
  • Alex notes that AI models like Grok, Gemini, ChatGPT, and Claude, when asked about the released UAP data, generally conclude they are normal phenomena or secret U.S. missions, not alien.
Also from this episode: (8)

AI Infrastructure (3)

  • Peter Diamandis and Dave Blundon highlight that AI demand is outpacing supply, leading to saturated chip markets and potential rate hikes for tokens, similar to the historical expansion of electricity uses.
  • Google is partnering with Planet Labs to build 'Project Suncatcher,' orbital data centers equipped with Tensor chips, with future plans likely involving Starship launches for mass deployment.
  • Leopold Aschenbrenner, fired from OpenAI's alignment team, now manages a $5.5 billion fund, leveraging 'situational awareness' to invest in critical AI infrastructure like chips and data centers, as highlighted by Dave Blundon.

Safety (1)

  • Peter launched the Future Vision XPRIZE, a global competition with $3.5 million in prize money, asking teams to create film trailers and treatments depicting hopeful, optimistic visions of the future to help align AI.

Agents (1)

  • Hermes Agent has surpassed OpenClaw as the number one on OpenRouter token ranking, offering a Python-based, more flexible, and recursively self-improving alternative for AI agents.

Chips (2)

  • Elon Musk's TerraFab project aims to produce 50x the current global chip production, with a potential cost as high as $119 billion, as he seeks to overcome chip supply bottlenecks.
  • Dave warns that if Taiwan's chip production, which supplies two-thirds of global GPUs, is disrupted by geopolitical events, everything discussed about AI development would grind to a halt.

Reasoning (1)

  • Alex argues the singularity will first be visible in space due to less incumbency and fewer regulatory hurdles compared to Earth, which has numerous entrenched interests and preservationist instincts.
#AI Infrastructure#Big Tech#Agents#Startups

Related Stories