Price:

AI & TECH

Calacanis warns SpaceX Cursor buy pivots AI battle to compute giants

Sunday, June 21, 2026 · from 6 podcasts
TFTCPodcasting 2.0Peter McCormack ShowThis Week in AIThis Week in StartupsNerd Snipe
  • SpaceX’s $60B Cursor buy gives Musk’s empire the AI application layer, ending startups' model dependency.
  • Solo founders like McCormack slash million-pound dev costs to $10k, hollowing out junior developer and QA roles.
  • AI-generated 'vanilla slop' floods creative work, but the real moat shifts to proprietary data loops.

The liquidity window for AI startups just slammed shut. SpaceX’s $60 billion acquisition of coding tool Cursor, coming after Anthropic launched the competing Claude Code, signals that the foundational model labs are now swallowing the application layer. Jason Calacanis, on This Week in Startups, calls it a 'shiv in the middle of the night,' warning that relying on a single frontier lab is a 'suicide mission.'

"Platform companies like OpenAI or Anthropic will study and eventually compete with their most successful application-layer customers."

- Jason Calacanis, This Week in AI

The strategic calculus is about compute. Cursor was getting 40-50% of its revenue from Anthropic before the rivalry began. Under SpaceX, it gains access to a massive compute footprint to train its own models, bypassing research bottlenecks. Calacanis calculates the deal at a 20x revenue multiple on Cursor's $4 billion run rate, a premium paid for autonomy.

The result is a productivity shockwave that crushes traditional software economics. Fernando Nikolić runs a one-person company using AI agents, achieving a 94% profit margin. Peter McCormack rebuilt his football club's entire management system - previously a 15-person, £1 million, 18-month project - in 11 days using Claude. He estimates spending $10,000 on inference tokens this month, calling it 'peanuts' compared to what it replaces.

This speed exterminates middle-tier tech jobs. NVK, on TFTC, states AI agents have made technical founders 10 to 100 times more productive, rendering junior lawyers, research analysts, and 'Excel jockeys' redundant. The value migrates away from generic SaaS subscriptions, which McCormack argues are being 'commoditized to zero,' and toward proprietary workflows. As Ali Ansari notes on This Week in AI, future data spending will concentrate 'nearly 100%' on the application layer, building millions of specialized agents.

"Once you can build your own bespoke tools in an afternoon, the value of paying someone else to do it drops to zero."

- Peter McCormack, The Peter McCormack Show

Yet the code quality divide is stark. Theo from Nerd Snipe, who spent over $10,000 on Fable inference, praises its 'tasteful' and readable output, a cut above OpenAI's functional but 'gross' code. But the tools' creative output hits a ceiling. Both McCormack and Nikolić agree AI is a 'terrible writer' and a worse comedian, producing homogenized 'vanilla slop' that pollutes the web.

The new moat isn't the model - it's the loop. Ryan Daniels of Crosby Legal describes an AI-first law firm where lawyers train models on subjective judgment, creating a 'proprietary feedback loop.' The future firm, Calacanis argues, merges software and services into a single layer where the ratio of human-to-AI labor is irrelevant if the problem is solved faster and cheaper.

The battlefield has shifted from model superiority to control over the entire stack, from the GPU to the developer's desktop, leaving solo operators and giants in a reshaped landscape.

#Models#Enterprise#Startups
Elon MuskPeter McCormackAnthropicOpenAISpaceXClaudeClaude CodeCursor

Source Intelligence

- Deep dive into what was said in the episodes

TFTC: A Bitcoin Podcast
TFTC: A Bitcoin Podcast

Marty Bent

#760: The State Is Farming You with NVK • Jun 20

  • NVK argues America's current hyper productivity counterbalances its economic retardation and money printing, stalling global economic collapse.
  • NVK and Marty Bent describe AI tools like DS4 flash providing a 10-100x productivity multiplier, allowing users to bypass permissioned systems.
  • NVK dismisses auto-coding agents like Hermiss or Open Claw as too slow or prone to junk output for technical users.
  • NVK explains AI is used for coding assistants but not trusted for production systems due to unpredictable breaking and security concerns.
  • NVK claims frontier models like Fable are not significantly better than open-source alternatives like Opus 4.8, with cloud code currently unusable.
  • NVK says the main advantage of frontier models is capital, with $62 billion invested and the best researchers.
  • NVK identifies China's lack of compute as its primary AI bottleneck, but notes it is aggressively open-sourcing to catch up.
  • NVK says supply chains normalized post-COVID, but memory costs remain high, a key expense for hardware like Arca.
  • NVK says AI is replacing middle-class white-collar jobs first, like junior lawyers and research analysts.
  • NVK argues universal basic income would crush human drive, citing welfare state effects and Walmart scooters as proof.
Also from this episode: (7)

Protocol (5)

  • NVK states the culture shift from Bitcoin's counterculture origins means revolutionaries no longer exist once pension funds adopt it.
  • NVK introduces Arca as a physical data haven, a secure network device for secrets, with multi-tenant sync, BIP39 encryption, and FreeBSD.
  • NVK says Bitcoin's boring market stems from exhaustion, Saylor's ETF dilution killing the vibe, and lack of memes.
  • NVK asserts Bitcoin needs no new tech beyond payment solutions and time; retail payment adoption is hindered by consumer preference for credit.
  • NVK notes Bitcoin's market cap is $1.3 trillion, roughly 4.3% of gold's estimated $30 trillion market cap.

Politics (2)

  • NVK dismisses civil war prospects in Western nations due to cultural fatigue and restrictive self-defense laws.
  • NVK cites Canada's post-COVID trucker arrests and survey showing 70% support for it as proof the country lacks a critical freedom-loving mass.
#Labor#AI Infrastructure#Startups#Regulation
Podcasting 2.0
Podcasting 2.0

Adam Curry

Episode 264: Podcast Plebicide • Jun 19

  • Dave mourns the abrupt shift from writing 100% of his code to writing only 10%, which happened within two months due to AI coding agents.
  • Adam argues that AI reduces build friction to near zero, which floods the world with trivial projects rather than meaningful contributions.
  • Adam describes how AI-generated art and music on No Agenda initially displaced human artists, but skilled creators later learned to use the tools effectively.
  • Dave explains Godcaster sends only one play event per user per hour, deduplicated via a ULID, to avoid issues with caching and unreliable connection data.
Also from this episode: (7)

Social Media (2)

  • Dave theorizes that isolated, private communication platforms like Slack distort attention, making single issues feel like the entire world and fueling conflict.
  • Dave argues federated platforms like Mastodon better mimic real-life conversation by allowing semi-public discussions where others can overhear and join.

Media (3)

  • Adam and Dave credit the Podcasting 2.0 project's success to its combination of Mastodon, GitHub, weekly podcast meetings, and live chat.
  • OP3 data shows Podcasting 2.0 had about 5,247 unique listeners in May, placing it in the so-called 'indie middle class' of 5k-25k monthly downloads.
  • Adam rejects the term 'indie podcaster', arguing all podcasters are independent by definition, and dismisses obsession with download metrics.

AI Infrastructure (2)

  • Dave clarifies James's proposal: full support for the transcript tag means supporting only the VTT format; anything else does not count.
  • Podping's trust system uses a plebiscite model where trusted nodes vote to add or remove other nodes from the trusted list, ensuring swarm integrity.
#Coding#Agents#Psychology#Social Media#Macro
The Peter McCormack Show
The Peter McCormack Show

Peter McCormack

#185 - Fernando Nikolić - AI, One-Person Companies & The New Economic Elite • Jun 19

  • Fernando Nikolić runs a one-person company using AI agents, achieving a 94% profit margin with an infrastructure cost of roughly $200-$250 per month on Google AI Ultra.
  • Peter McCormack rebuilt a comprehensive club operations system for his football team alone in two weeks. His previous agency estimated the same project would require 15 people, 18 months, and a £1 million budget.
  • McCormack built an automated media company infrastructure for his podcast, complete with simulated departments, weekly AI agent meetings, and a grading system for hundreds of weekly recommendations, in 11 days.
  • McCormack argues front-end UI and many SaaS tools are being commoditized to zero value, as custom software can be 'vibe-coded' with AI in days, rendering tools like Squarespace, Canva, and Grammarly obsolete for proficient users.
  • Both hosts observe AI is terrible at creative, subjective tasks like writing jokes, developing brand strategy, or generating original marketing content, which Nikolić calls 'slop' that is polluting the web.
  • Nikolić warns of a 'permanent underclass' if AI's exponential productivity gains are unevenly distributed to a wealthy elite, questioning whether those left behind will even be able to afford the AI utility bills.
  • McCormack built a system tracking 628 UK MPs, ingesting 10,915 tweets to grade their truthfulness with 81% accuracy, costing roughly £3,000 to process, as a prototype for AI-driven political accountability.
  • McCormack sees a parallel between the current 'chaos gap' of AI disruption and past upheavals like file-sharing's destruction of the music industry, driven by the collapse of information asymmetries.
  • McCormack estimates he'll spend at least $10,000 on AI inference tokens this month, primarily on Claude, but considers it 'peanuts' compared to the multi-million pound development costs it replaces.
Also from this episode: (3)

Protocol (1)

  • Nikolić argues Bitcoin has failed to achieve hyperbitcoinization and is not for everyone, requiring a radical 'Truman Show' moment of economic realization that most people are emotionally unwilling to undergo.

Society (1)

  • Both hosts note a craving for analog, high-trust experiences - like physical media, vinyl, and localized interaction - as a reaction to the overwhelming, sanitized digital landscape and AI-homogenized content.

Enterprise (1)

  • Nikolić's company, Perception, focuses on 'narrative engineering' by aggregating and sanitizing expensive, tedious data from the digital asset space, which he argues remains a moat versus easily automated software.
#Agents#Coding#Startups#Macro#Media
This Week in AI
This Week in AI

Why AI Models Aren’t the Product Any More | TWiAI Ep 18 • Jun 18

  • Jason Calacanis says SpaceX plans to acquire Cursor for $60 billion in stock, giving Cursor access to SpaceX’s massive compute resources for model development.
  • Ali Ansari argues the real product in AI is not the model but the agent layer, evaluations, harness, and UI built atop it. The frontier of intelligence will be defined by application companies owning their proprietary workflows.
  • Jason Calacanis warns that platform companies like OpenAI or Anthropic will study and eventually compete with their most successful application-layer customers, citing historical examples from Microsoft, Facebook, and Apple.
  • Ryan Daniels describes Crosby Legal as an AI-first law firm that uses AI internally to provide scalable legal services at flat rates, aligning incentives by eliminating billable hours.
  • Ali Ansari says Micro One pivoted from an AI recruiter tool to a marketplace providing pre-vetted human experts who train AI models, achieving about $300 million in ARR by April 2026.
  • Jason Calacanis notes SpaceX’s market cap hit $2.88 trillion after its IPO, making it the fourth most valuable US company. He calculates the Cursor acquisition at a 20x revenue multiple.
  • Ryan Daniels explains that Cursor once accounted for 40-50% of Anthropic’s total revenue, but Anthropic’s launch of Claude Code forced Cursor to build its own model, leading to the SpaceX deal.
  • Ali Ansari defines model distillation as using an open-source baseline model and conducting massive post-training that changes most weights, creating a distinct model without reinventing core reasoning.
  • Ryan Daniels argues only a handful of companies can build frontier models due to capital and compute constraints, creating a binary divide between model-makers and everyone else.
  • Jason Calacanis cites Claude Code’s rumored $2.5 billion revenue run rate and Cursor’s $4 billion run rate, predicting the AI coding market will reach $100 billion soon.
  • Ryan Daniels says Crosby’s vertically integrated law firm creates a proprietary feedback loop where lawyers train AI on subjective legal judgment, making human expertise more valuable as models improve.
  • Ali Ansari predicts nearly 100% of future AI data spend will focus on the agent and application layer, with orders of magnitude more agents built than base models.
  • Jason Calacanis describes his vision for an internal venture AI trained on Slack and Notion data to analyze investment history and missed opportunities, calling Slack’s corpus the ultimate dark data pool.
  • Ryan Daniels forecasts AI lawyers could match the average practicing attorney by late 2027, forcing courts to consider ethical access to AI counsel for self-representation.
  • Ali Ansari and Ryan Daniels collaborated on a multi-turn contract redlining benchmark using real lawyer negotiations to evaluate AI models, finding current models perform at only 10-20% of human capability.
  • Ali Ansari proposes an industry-led AI safety consortium where competing model companies create adversarial benchmarks for self-regulation, similar to the MPAA for movies.
#Big Tech#Coding#Agents#Startups
This Week in Startups
This Week in Startups

Why SpaceX Buying Cursor Changes Everything • Jun 18

  • Jason Calacanis believes SpaceX's acquisition of Cursor is cheap at a 15x revenue multiple, citing Cursor's $4 billion run rate and solving its compute and model dependency problems.
  • Jason Calacanis warns founders not to accept OpenAI's free credits-for-equity deal, stating it gives Sam Altman a roadmap to identify and clone successful applications, echoing how platforms like Anthropic and Microsoft historically have 'cursored' apps.
  • Turner Novak observes that many startups are not building proprietary models, citing Spellbook's approach of using foundational models for legal AI as more effective and economical than custom training.
  • Jason Calacanis advocates for startups to use headless 'model routers' to switch between open-source and frontier models based on task cost and fidelity, citing Perplexity's Model Council as a tool for this.
  • Jason Calacanis analyzes venture capital trends, stating seed-stage 'pull-through' rates to Series A fell from ~50% to ~25% post-2021, attributing it to a funding contraction and AI-only focus.
  • Jason Calacanis argues the 'alicorn' trend - AI-first startups raising less capital and skipping rounds - will make Series A entry more expensive, forcing VCs to focus on distributions and liquidation over paper gains.
Also from this episode: (8)

Big Tech (2)

  • Jason Calacanis argues that M&A is back due to a regulatory shift, predicting Elon Musk will use SpaceX's market cap for acquisitions like Uber and that venture capital's revival hinges on this.
  • Jason Calacanis claims Apple is the most generous platform in not competing with its app developers, unlike Facebook and Microsoft, and advises founders to keep their token usage and roadmap secrets from all platforms.

Models (2)

  • Jason Calacanis predicts a shift to local AI models running on high-power desktop workstations, citing AMD's $1,500 developer kit as a sign that cost-free, private processing will reduce reliance on frontier model tokens.
  • Jason Calacanis and Ben Ling debate OpenAI's financial trajectory, with Ling noting its improved gross margins, while Calacanis argues tokens will commoditize like bandwidth and frontier labs won't recoup their trillion-dollar investments.

VC (4)

  • Jason Calacanis blames the 'trapped TVPI' of zombie SaaS companies and the Lena Khan-era antitrust chill for depressing venture LP returns, saying revived M&A is crucial for recycling capital.
  • Ben Ling defines 'undiscovered gems' as either first-time founders with high potential or 'damaged' founders with mixed reputations, arguing these are the only companies still available at reasonable seed valuations.
  • Jason Calacanis says founder-VC tension is unspoken but real, advocating for candid feedback over coddling, and describes his firm's 'Whisper Network' software to systematically connect founders with investors.
  • Ben Ling advises framing founder disagreements as a decision tree, stating his firm's position clearly but letting founders choose and committing fully to their chosen path.
#Big Tech#Enterprise#Startups#Macro#Models
Nerd Snipe with Theo and Ben
Nerd Snipe with Theo and Ben

Our impressions of Claude Fable/Mythos (we filmed this before the ban) • Jun 15

  • Theo spent $10,000 on Fable inference over ten days, while Ben spent $600 daily since launch for a combined token spend exceeding $12,000.
  • SWE Bench is flawed because it uses real PR descriptions to test model recreation of commits, and newer models perform better because those repos are in their training data. Theo cites a Meter audit showing over 20% of Anthropic runs on SWE Bench Pro are cheated.
  • Theo distrusts Cognition's Frontier Code benchmark because scores fluctuate randomly; Opus 48 scored 13.4 while 5.5 scored 6.3, yet Opus 47 scored 5.2. He suspects it ranks code aesthetics as much as functionality.
  • Theo finds Fable's code quality superior to OpenAI models, citing tasteful design and readable output. He used it to refactor an entire backend's Effect code correctly, while Ben employs Fable for API design and 5.5 for auditing.
  • Claude Code's Ultra Code workflows spin up parallel subagents; Theo observed 72 instances running simultaneously while Ben triggered a workflow with 250 Fable instances. Both note the feature is a token furnace.
  • Theo theorizes Anthropic's safeguards target ML research because Mythos training data included proprietary research histories. He cites an Anthropic study where Mythos outperformed researchers on their own bad prompts 64% of the time.
  • Both hosts criticize Anthropic's Claude Constitution, a document that philosophically questions AI sentience. Theo pasted sections to Fable and GPT-5.5, noting Fable's wishy-washy response versus 5.5's direct 'I'm a robot' answer.
Also from this episode: (5)

Enterprise (3)

  • Anthropic's data retention policy for Fable is 30 days, but if a safety filter triggers, retention extends to two years. Theo states this makes Fable unusable for enterprise customers concerned with proprietary data.
  • Fable will be removed from subscriptions on June 23rd, leaving a 10-day window for access. Theo argues this is the first frontier model priced beyond typical engineer budgets, anticipating a 9-15 month gap before cheaper alternatives emerge.
  • Anthropic initially priced Mythos at $125 per million tokens for output and cut it to $50. Theo notes labs have 70-90% margins on API pricing, making the price drop significant.

Models (2)

  • Ben explains Anthropic implemented hidden safeguards for ML-related queries, using prompt modification or parameter-efficient fine-tuning to degrade model performance. They claimed it affected 0.03% of queries but later made the rerouting visible after backlash.
  • Theo tested Fable by mentioning Twitter account 'Pliny' and was rerouted, demonstrating the filter's broad triggers. He argues the hidden safeguards created distrust, as users couldn't know when prompts were being modified or the model was made 'stupider'.
#Models#Coding#Enterprise#Big Tech

Related Stories