{
 "episode_id": "all-in-2026-07-15-former-intel-ceo-on-what-went-wrong-what-s-next-lovable-ceo-",
 "title": "Former Intel CEO on What Went Wrong, What's Next + Lovable CEO on the Real Promise of Vibe Coding",
 "podcast": "all-in",
 "published_at": "2026-07-15",
 "speakers": [
  "Jason Calacanis",
  "Pat Gelsinger",
  "Anton Osika"
 ],
 "summary": "In a two-part interview taped in Paris, Jason Calacanis first sits with former Intel CEO Pat Gelsinger for a post-mortem on Intel's decline: Gelsinger argues the company lost its way once business and finance people rather than technologists ran it, returned about $100B to shareholders instead of building fabs and buying EUV machines, and missed the shifts to Apple Silicon, NVIDIA's CUDA-powered GPUs, and TSMC's foundry model. Gelsinger then lays out a macro thesis — Taiwan's fab concentration and sub-three-week energy reserves are a civilization-scale supply-chain risk; energy capacity is the natural governor that keeps the AI buildout from becoming a runaway bubble; token value as a measure of intelligence is 'somewhat infinite'; and quantum computing will deliver meaningful results before 2030 (with 'Q-day' encryption breaks by roughly 2033), converging with classical and AI compute in a 'trinity of computing.' In the second half, Lovable founder Anton Osika describes a platform seeing a million new products a week, 700M+ monthly app visits, and $500M in revenue after 20 months, serving an roughly 80% non-technical customer base by imposing opinionated architecture, security, and payments so builders 'don't go off a cliff.' Osika argues vibe coding has moved from mockups to deployable, secure production software, that bespoke apps will increasingly replace generic SaaS while interoperating with it, and that Lovable improves its own open-weight models via post-training and reinforcement learning on the mistakes frontier models make across its enormous token distribution. Both close on the human bottleneck: even with more capable models like Anthropic's Fable, deciding what to build — not writing code — is the constraint.",
 "topics": [
  {
   "name": "Intel's decline: technologists vs. business leaders",
   "timestamp": "1:41",
   "summary": "Gelsinger argues Intel went off the rails once business and finance people rather than technologists ran it, and that returning ~$100B to shareholders via dividends and buybacks starved it of capital for new fabs and EUV machines."
  },
  {
   "name": "Apple Silicon and Steve Jobs' long game",
   "timestamp": "5:33",
   "summary": "Gelsinger recounts how Steve Jobs quietly built Apple's internal silicon competence over multiple OS releases to move off Intel and co-optimize system and chip design."
  },
  {
   "name": "NVIDIA, CUDA, and GPUs becoming general-purpose compute",
   "timestamp": "7:56",
   "summary": "Gelsinger describes how Intel scoffed at Jensen Huang's GPUs until CUDA and HPC turned them into general-purpose compute, and how Intel's competing Larrabee project was killed a week after his first departure."
  },
  {
   "name": "TSMC and the rise of the foundry model",
   "timestamp": "11:32",
   "summary": "Gelsinger explains how TSMC's foundry vision beat Intel's integrated IDM model to become the semiconductor industry standard, out-producing Intel 5-7x on wafers, and why Intel's comeback requires becoming a foundry."
  },
  {
   "name": "Onshoring chips, the CHIPS Act, and Taiwan blockade risk",
   "timestamp": "14:11",
   "summary": "Gelsinger credits the CHIPS Act with raising US leading-edge share from ~12% to ~18%, then warns that Taiwan's sub-three-week energy reserves make a Chinese blockade a brownout risk with economic impact greater than the Great Depression."
  },
  {
   "name": "Is AI a bubble? Energy as the governor and token economics",
   "timestamp": "17:28",
   "summary": "Gelsinger argues energy capacity is a hard upper bound that keeps the AI buildout from overheating, and that the value of a token as a measure of intelligence is 'somewhat infinite,' targeting a 10,000x reduction in cost and energy per token."
  },
  {
   "name": "AI valuations, healthy corrections, and the trinity of computing",
   "timestamp": "21:21",
   "summary": "Gelsinger contrasts today's real-revenue AI companies with dot-com speculation, welcomes periodic corrections, and points to the convergence of classical, AI, and quantum computing — predicting meaningful quantum results before 2030."
  },
  {
   "name": "Lovable: mission, growth, and the customer for vibe coding",
   "timestamp": "25:30",
   "summary": "Anton Osika frames Lovable's mission as empowering humans, citing a million new products a week, 700M+ monthly app visits, 50M+ apps, 20 months in market, and a customer base that is roughly 80% non-technical."
  },
  {
   "name": "From mockups to production software: the intranet story",
   "timestamp": "29:17",
   "summary": "Jason recounts an employee building a full Founder University intranet on Lovable in about four hours for under $2,000 versus a ~$500K legacy cost, while Osika describes Lovable handling architecture, payments, and security for non-technical builders."
  },
  {
   "name": "Bespoke software replacing SaaS incumbents",
   "timestamp": "36:29",
   "summary": "Osika's Nursa story (replacing 10+ internal tools, saving $1M+/yr) and his view that bespoke apps will grow for specific requirements while Lovable keeps interoperating with Salesforce, HubSpot, Slack, Google, and Microsoft."
  },
  {
   "name": "Lovable's model strategy, token economics, co-opetition, and Fable",
   "timestamp": "39:40",
   "summary": "Osika describes multi-model routing toward open weights, a Stockholm post-training/RL team mining frontier-model mistakes, caps with overages (~60% of low-tier users top up), the CERN 'co-opetition' model, and his read on Anthropic's Fable."
  }
 ],
 "entities": [
  {
   "name": "Jason Calacanis",
   "type": "person",
   "salience": "central",
   "role_in_episode": "Host and interviewer for both the Pat Gelsinger (Intel/AI/quantum) and Anton Osika (Lovable) conversations"
  },
  {
   "name": "Pat Gelsinger",
   "type": "person",
   "salience": "central",
   "role_in_episode": "Former Intel CEO, now an investor; delivers the Intel post-mortem and his AI-buildout, energy, and quantum thesis"
  },
  {
   "name": "Anton Osika",
   "type": "person",
   "salience": "central",
   "role_in_episode": "Founder and CEO of Lovable; explains its growth, customer base, model strategy, and product roadmap"
  },
  {
   "name": "Steve Jobs",
   "type": "person",
   "salience": "discussed",
   "role_in_episode": "Apple co-founder; Gelsinger recounts how Jobs quietly built Apple's silicon competence to move off Intel"
  },
  {
   "name": "Jensen Huang",
   "type": "person",
   "salience": "discussed",
   "role_in_episode": "NVIDIA CEO; Gelsinger describes how his GPUs plus CUDA became general-purpose compute Intel underestimated"
  },
  {
   "name": "Andy Grove",
   "type": "person",
   "salience": "mentioned",
   "role_in_episode": "Intel technical leader and Gelsinger's mentor, cited as the archetype of Intel's engineering culture"
  },
  {
   "name": "Gordon Moore",
   "type": "person",
   "salience": "mentioned",
   "role_in_episode": "Intel co-founder cited among the deeply technical leaders Gelsinger grew up under"
  },
  {
   "name": "Bob Noyce",
   "type": "person",
   "salience": "mentioned",
   "role_in_episode": "Intel co-founder cited among Gelsinger's technical mentors"
  },
  {
   "name": "Satya Nadella",
   "type": "person",
   "salience": "mentioned",
   "role_in_episode": "Cited (as 'Sacha') as a deeply technical non-founder CEO supporting Gelsinger's technical-leadership thesis"
  },
  {
   "name": "Sundar Pichai",
   "type": "person",
   "salience": "mentioned",
   "role_in_episode": "Cited (as 'Sundar') as a deeply technical non-founder CEO"
  },
  {
   "name": "Nenad",
   "type": "person",
   "salience": "mentioned",
   "role_in_episode": "Nursa employee who used Lovable to build Nursa Study and replace 10+ internal tools"
  },
  {
   "name": "Intel",
   "type": "company",
   "salience": "central",
   "role_in_episode": "Subject of Gelsinger's post-mortem on how business leadership, missed fabs, and buybacks cost its lead"
  },
  {
   "name": "Apple",
   "type": "company",
   "salience": "discussed",
   "role_in_episode": "Built its own silicon to leave Intel; also cited for spending its capital hoard on buybacks and small acquisitions"
  },
  {
   "name": "NVIDIA",
   "type": "company",
   "salience": "discussed",
   "role_in_episode": "Its GPUs plus CUDA became general-purpose AI compute; the platform Intel scoffed at"
  },
  {
   "name": "TSMC",
   "type": "company",
   "salience": "central",
   "role_in_episode": "Pioneered the foundry model that became the semiconductor industry standard, out-producing Intel 5-7x on wafers"
  },
  {
   "name": "Samsung",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Cited as another foundry whose factories are operating at scale"
  },
  {
   "name": "Lovable",
   "type": "company",
   "salience": "central",
   "role_in_episode": "Anton Osika's vibe-coding platform; its growth, economics, and roadmap anchor the episode's second half"
  },
  {
   "name": "Nursa",
   "type": "company",
   "salience": "discussed",
   "role_in_episode": "US company where Nenad used Lovable to launch Nursa Study and replace 10+ internal tools, saving $1M+/yr"
  },
  {
   "name": "AWS",
   "type": "company",
   "salience": "discussed",
   "role_in_episode": "Both a Lovable hosting partner 'under the hood' and the incumbent its new hosting product competes with"
  },
  {
   "name": "Anthropic",
   "type": "company",
   "salience": "discussed",
   "role_in_episode": "Maker of the Fable model Anton evaluates; a frontier-model partner and competitor to Lovable"
  },
  {
   "name": "CERN",
   "type": "organization",
   "salience": "discussed",
   "role_in_episode": "Where Anton worked; source of the 'co-opetition' parallel-team model he now applies at Lovable"
  },
  {
   "name": "Cursor",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "One of the vibe-coding tools Jason's team tried before an employee settled on Lovable"
  },
  {
   "name": "Perplexity",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Its 'Computer' co-work tool cited among the tools Jason's team experimented with"
  },
  {
   "name": "Salesforce",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Incumbent SaaS cited in the bespoke-vs-SaaS debate; Lovable interoperates with it"
  },
  {
   "name": "HubSpot",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Incumbent SaaS cited in the bespoke-vs-SaaS debate"
  },
  {
   "name": "Google",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Its Google Suite cited as incumbent software Lovable connects to under the hood"
  },
  {
   "name": "Microsoft",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Its Office/Suite cited as incumbent software Lovable interoperates with"
  },
  {
   "name": "Slack",
   "type": "product",
   "salience": "mentioned",
   "role_in_episode": "SaaS tool Jason muses about replacing with bespoke vibe-coded software"
  },
  {
   "name": "Cerebras",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Inference-hardware company (rendered 'Cerebris') Jason cites as making inference cheap"
  },
  {
   "name": "Groq",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Inference company (rendered 'Grok') cited alongside Cerebras"
  },
  {
   "name": "Apple Silicon",
   "type": "product",
   "salience": "discussed",
   "role_in_episode": "Apple's in-house chips; the outcome of Jobs' long game to leave Intel"
  },
  {
   "name": "CUDA",
   "type": "product",
   "salience": "discussed",
   "role_in_episode": "NVIDIA's software stack that turned GPUs into general-purpose compute for HPC and AI"
  },
  {
   "name": "Fable",
   "type": "product",
   "salience": "discussed",
   "role_in_episode": "Anthropic's latest model; Anton reports strong first-attempt output but no shift in the human planning bottleneck"
  },
  {
   "name": "Larrabee",
   "type": "product",
   "salience": "mentioned",
   "role_in_episode": "Intel's GPU-style x86 project (rendered 'Larabee'), killed a week after Gelsinger's first departure"
  },
  {
   "name": "Pentium",
   "type": "product",
   "salience": "mentioned",
   "role_in_episode": "The Intel chip behind the 'Intel inside' era"
  },
  {
   "name": "Centrino",
   "type": "product",
   "salience": "mentioned",
   "role_in_episode": "Intel chip Apple adopted before building its own silicon"
  },
  {
   "name": "PsiQuantum",
   "type": "company",
   "salience": "mentioned",
   "role_in_episode": "Gelsinger's portfolio quantum-computing company and his pick in the quantum race"
  },
  {
   "name": "Fortnite",
   "type": "product",
   "salience": "mentioned",
   "role_in_episode": "Cited as an example of a 3D game built with Anthropic's Fable"
  },
  {
   "name": "Founder University",
   "type": "organization",
   "salience": "discussed",
   "role_in_episode": "Jason's pre-accelerator; an employee built its entire intranet on Lovable in about four hours"
  },
  {
   "name": "Nursa Study",
   "type": "product",
   "salience": "mentioned",
   "role_in_episode": "Nursa's nurse-education product line built on Lovable"
  },
  {
   "name": "semiconductors",
   "type": "concept",
   "salience": "central",
   "role_in_episode": "The industry through which Gelsinger frames Intel's decline, TSMC's rise, and Taiwan risk"
  },
  {
   "name": "foundry model",
   "type": "concept",
   "salience": "discussed",
   "role_in_episode": "TSMC's third-party manufacturing model that displaced Intel's integrated IDM approach"
  },
  {
   "name": "CHIPS Act",
   "type": "concept",
   "salience": "discussed",
   "role_in_episode": "US onshoring policy Gelsinger credits with raising US leading-edge share from ~12% to ~18%"
  },
  {
   "name": "AI bubble",
   "type": "concept",
   "salience": "central",
   "role_in_episode": "The core debate of the Gelsinger segment: whether the AI buildout is overheating"
  },
  {
   "name": "AI buildout",
   "type": "concept",
   "salience": "central",
   "role_in_episode": "The multi-decade data-center and chip expansion Gelsinger sees energy capacity as the governor of"
  },
  {
   "name": "token economics",
   "type": "concept",
   "salience": "central",
   "role_in_episode": "Cost and energy per token and its near-infinite value as a measure of intelligence"
  },
  {
   "name": "quantum computing",
   "type": "concept",
   "salience": "central",
   "role_in_episode": "Gelsinger predicts meaningful results before 2030 and 'Q-day' encryption breaks by roughly 2033"
  },
  {
   "name": "vibe coding",
   "type": "concept",
   "salience": "central",
   "role_in_episode": "Lovable's category; building deployable software directly instead of mockups"
  },
  {
   "name": "bespoke software",
   "type": "concept",
   "salience": "central",
   "role_in_episode": "Custom vibe-coded apps potentially replacing generic SaaS incumbents"
  },
  {
   "name": "open-weight models",
   "type": "concept",
   "salience": "discussed",
   "role_in_episode": "A growing share of the models Lovable routes to under its multi-model strategy"
  },
  {
   "name": "post-training",
   "type": "concept",
   "salience": "discussed",
   "role_in_episode": "What Lovable's Stockholm research team does to improve its own models"
  },
  {
   "name": "reinforcement learning",
   "type": "concept",
   "salience": "mentioned",
   "role_in_episode": "Method Lovable applies to the specific failures of frontier models on its token distribution"
  },
  {
   "name": "co-opetition",
   "type": "concept",
   "salience": "discussed",
   "role_in_episode": "The CERN parallel-team model Anton champions for rapid experimentation"
  },
  {
   "name": "Jevons paradox",
   "type": "concept",
   "salience": "discussed",
   "role_in_episode": "Cited by both: cheaper, better tokens drive explosive usage growth"
  },
  {
   "name": "Taiwan",
   "type": "place",
   "salience": "central",
   "role_in_episode": "Its fab concentration and sub-three-week energy reserves make it the episode's key geopolitical risk"
  },
  {
   "name": "China",
   "type": "place",
   "salience": "discussed",
   "role_in_episode": "Cited as having blockaded the Taiwan Strait seven times in four years"
  },
  {
   "name": "United States",
   "type": "place",
   "salience": "discussed",
   "role_in_episode": "Backdrop for onshoring, the CHIPS Act, and self-hosting AI on US soil"
  },
  {
   "name": "Stockholm",
   "type": "place",
   "salience": "mentioned",
   "role_in_episode": "Home of Lovable's post-training research team, part of the European AI ecosystem"
  }
 ],
 "claims": [
  {
   "holder": "Pat Gelsinger",
   "claim": "Intel's fundamental failure was being run by business and finance people rather than technologists, since business leaders promote other business leaders and hollow out technical decision-making.",
   "stance": "opinion",
   "conviction": 0.85,
   "subjects": [
    "Intel"
   ],
   "timestamp": "2:08",
   "quote": "one of the things that went off the rail was when it started to be run by business people."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "Hardcore technical decisions that affect billions of dollars cannot be made through a spreadsheet, so a technology business fundamentally needs technologists running it.",
   "stance": "opinion",
   "conviction": 0.85,
   "subjects": [
    "Intel",
    "semiconductors"
   ],
   "timestamp": "3:01",
   "quote": "when you're making these hardcore technical decisions that affect billions of dollars, you don't do that through a spreadsheet."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "Returning ~$100B to shareholders in the years before he came back was a mistake; that capital should have stayed on Intel's balance sheet to build fabs and buy EUV machines.",
   "stance": "opinion",
   "conviction": 0.8,
   "subjects": [
    "Intel"
   ],
   "timestamp": "3:40",
   "quote": "What I wouldn't have done for another $100 billion on the balance sheet."
  },
  {
   "holder": "Jason Calacanis",
   "claim": "Apple squandered the opportunity of its enormous capital hoard by only doing small, marginal acquisitions instead of big transformative bets.",
   "stance": "opinion",
   "conviction": 0.6,
   "subjects": [
    "Apple"
   ],
   "timestamp": "4:37",
   "quote": "They buy little, tiny acquisitions on the margins."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "Steve Jobs deliberately built Apple's internal semiconductor competence over years to reduce reliance on Intel and co-optimize system and silicon design.",
   "stance": "assertion",
   "conviction": 0.8,
   "subjects": [
    "Apple",
    "Steve Jobs",
    "Apple Silicon"
   ],
   "timestamp": "5:33",
   "quote": "He had been preparing the core technologies inside of Apple for something that might happen in the future."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "Killing Intel's Larrabee project — an attempt to make x86 do what NVIDIA's GPUs did — was a pivotal error; had it survived, the industry would look very different.",
   "stance": "opinion",
   "conviction": 0.6,
   "subjects": [
    "Intel",
    "Larrabee",
    "NVIDIA"
   ],
   "timestamp": "8:37",
   "quote": "the project was killed a week after I left. And the world would have been so much different."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "TSMC's foundry model became the model of the entire semiconductor industry, displacing Intel's integrated design-and-manufacturing approach except for Intel itself and memory.",
   "stance": "assertion",
   "conviction": 0.85,
   "subjects": [
    "TSMC",
    "Intel",
    "foundry model"
   ],
   "timestamp": "11:32",
   "quote": "that model of foundry became the model of the semiconductor industry, with two exceptions, Intel and memory."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "A core thesis of Intel's comeback strategy is that Intel must become a foundry serving third parties, not just an integrated manufacturer for its own chips.",
   "stance": "opinion",
   "conviction": 0.8,
   "subjects": [
    "Intel",
    "foundry model"
   ],
   "timestamp": "11:32",
   "quote": "We must become a foundry as well."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "A Chinese blockade of Taiwan could brown out the island and its fabs without a shot being fired, with economic impact greater than the Great Depression.",
   "stance": "assertion",
   "conviction": 0.8,
   "subjects": [
    "Taiwan",
    "semiconductors",
    "China"
   ],
   "timestamp": "15:13",
   "quote": "The economic impact of a brownout of Taiwan is greater than the Great Depression in the world."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "The US needs far more resilient semiconductor supply chains because China's intentions toward Taiwan have been clear over a sustained period.",
   "stance": "opinion",
   "conviction": 0.8,
   "subjects": [
    "Taiwan",
    "China",
    "semiconductors"
   ],
   "timestamp": "17:10",
   "quote": "Their intentions have been clear over a sustained period of time. We need more resilient supply chains for it."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "Energy capacity is a hard upper bound on the AI buildout — nobody builds data centers or buys GPUs without energy — which guarantees the AI bubble can't get too far ahead of itself.",
   "stance": "opinion",
   "conviction": 0.8,
   "subjects": [
    "AI bubble",
    "AI buildout"
   ],
   "timestamp": "18:28",
   "quote": "there is a silver lining here that guarantees we don't get too far ahead of ourselves in terms of bubble, and that is energy capacity."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "The value of a token as a measure of intelligence is 'somewhat infinite,' so the economic potential AI unleashes is vast, especially given developed-world labor shortages.",
   "stance": "opinion",
   "conviction": 0.75,
   "subjects": [
    "token economics"
   ],
   "timestamp": "18:41",
   "quote": "the potential value that we unleash in a token economic world is somewhat infinite, particularly with labor shortages"
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "There has never been a better time in human history to be a technologist or to be alive, because AI will let us solve chemistry, language, materials, cancer, and poverty.",
   "stance": "opinion",
   "conviction": 0.85,
   "subjects": [
    "AI buildout"
   ],
   "timestamp": "20:38",
   "quote": "There has not been a time in human history where it's been better to be a technologist than the one we're in right now."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "Unlike dot-com speculation, today's AI companies have real revenues and margins; periodic corrections that keep multiples reasonable are healthy because the trajectory won't be a smooth curve.",
   "stance": "opinion",
   "conviction": 0.75,
   "subjects": [
    "AI bubble"
   ],
   "timestamp": "21:45",
   "quote": "periodic corrections that keep the multiple, you know, earnings multiples and so on, and reasonable things is good because this will not be a smooth curve."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "The most exciting frontier is the 'trinity of computing' — classical, AI, and quantum computing converging.",
   "stance": "opinion",
   "conviction": 0.7,
   "subjects": [
    "quantum computing"
   ],
   "timestamp": "21:45",
   "quote": "what I call the trinity of computing, classical computing, AI computing, and quantum computing. And when those three come together, okay, that's when things get really exciting."
  },
  {
   "holder": "Pat Gelsinger",
   "claim": "Quantum error correction has now been proven across multiple modalities, so the remaining challenge is engineering scale, not the underlying physics.",
   "stance": "assertion",
   "conviction": 0.75,
   "subjects": [
    "quantum computing"
   ],
   "timestamp": "23:15",
   "quote": "We know how to build qubits. We know how to error correct qubits. We now have algorithmics against quantum, and now it's just about the engineering scale."
  },
  {
   "holder": "Jason Calacanis",
   "claim": "Vibe coding has leapt in a year from producing mockups and wireframes to building deployable, secure production software directly.",
   "stance": "assertion",
   "conviction": 0.8,
   "subjects": [
    "vibe coding",
    "Lovable"
   ],
   "timestamp": "34:53",
   "quote": "A year ago, we were at mockups. Now we're at functionality and secure and super viable for deployment."
  },
  {
   "holder": "Anton Osika",
   "claim": "Many engineers no longer read or write code, which means you no longer need to be an engineer to create software.",
   "stance": "assertion",
   "conviction": 0.8,
   "subjects": [
    "vibe coding",
    "Lovable"
   ],
   "timestamp": "30:09",
   "quote": "many engineers, they don't look at the code, they don't write code anymore, and that means that you don't need to be an engineer to create software."
  },
  {
   "holder": "Anton Osika",
   "claim": "Lovable's core value is imposing opinionated architecture and best practices — payments, security, discoverability — so non-technical builders don't go off a cliff.",
   "stance": "opinion",
   "conviction": 0.8,
   "subjects": [
    "Lovable",
    "vibe coding"
   ],
   "timestamp": "30:09",
   "quote": "it creates a structure for the architecture of the software that you build, and it makes sure that you don't go off a cliff"
  },
  {
   "holder": "Anton Osika",
   "claim": "Being opinionated from day one and focusing on the non-technical '99%' is what made Lovable successful.",
   "stance": "opinion",
   "conviction": 0.8,
   "subjects": [
    "Lovable"
   ],
   "timestamp": "30:09",
   "quote": "us being opinionated from day one and being focused on making this for the 99%."
  },
  {
   "holder": "Jason Calacanis",
   "claim": "Foundational SaaS like Slack could be replaced by bespoke vibe-coded software integrated more deeply into a company's operations.",
   "stance": "opinion",
   "conviction": 0.6,
   "subjects": [
    "bespoke software",
    "Slack"
   ],
   "timestamp": "36:56",
   "quote": "maybe I should vibe code my own Slack."
  },
  {
   "holder": "Anton Osika",
   "claim": "Where existing tools don't fit specific requirements, companies will increasingly build bespoke solutions, even as Lovable continues to interoperate with the incumbent SaaS tools underneath.",
   "stance": "opinion",
   "conviction": 0.7,
   "subjects": [
    "bespoke software",
    "Lovable"
   ],
   "timestamp": "37:46",
   "quote": "In those cases, I think yes, you will have more bespoke solutions."
  },
  {
   "holder": "Anton Osika",
   "claim": "Lovable's strategy is to always do whatever is best for the customer, which means routing every request to the most suitable model across multiple commercial and open-weight vendors.",
   "stance": "assertion",
   "conviction": 0.8,
   "subjects": [
    "Lovable",
    "open-weight models"
   ],
   "timestamp": "40:22",
   "quote": "we're using multiple models. And so if you ask Lovable now, it's actually routed to the model that's most suitable to whatever you want to do."
  },
  {
   "holder": "Anton Osika",
   "claim": "Lovable never picks a cheaper model when it is measurably worse for customers, because moving fast with more intelligence is usually worth the spend.",
   "stance": "opinion",
   "conviction": 0.75,
   "subjects": [
    "Lovable",
    "token economics"
   ],
   "timestamp": "43:23",
   "quote": "We've never had the decision to say, let's use a cheaper model here if it's measurably worse for our customers."
  },
  {
   "holder": "Anton Osika",
   "claim": "Lovable improves its own models by mining the specific mistakes frontier models make across its enormous token distribution and applying reinforcement learning to those failures.",
   "stance": "assertion",
   "conviction": 0.75,
   "subjects": [
    "Lovable",
    "post-training",
    "reinforcement learning"
   ],
   "timestamp": "42:06",
   "quote": "we're looking at the mistakes that any of the models do right now, and then we prioritize them by what drives most impact for our customers, and then we make the models, we create datasets, we did something called reinforcement learning specifically for the problems where the frontier models are making mistakes for us right now."
  },
  {
   "holder": "Anton Osika",
   "claim": "Now that engineering is no longer the bottleneck, running many isolated teams at the same problem ('co-opetition') is a good way to avoid local minima and find the best solution.",
   "stance": "opinion",
   "conviction": 0.75,
   "subjects": [
    "co-opetition",
    "vibe coding"
   ],
   "timestamp": "45:54",
   "quote": "now since the engineering is less of the bottleneck, it's more the question of what is the right thing to build. I think it's a great thing to have if you have the sufficiently many humans right to do to try to attempt the solving the same problem in different ways."
  },
  {
   "holder": "Anton Osika",
   "claim": "Anthropic's Fable can produce very sophisticated, good-looking output on the first attempt, especially 3D games.",
   "stance": "opinion",
   "conviction": 0.7,
   "subjects": [
    "Fable"
   ],
   "timestamp": "47:34",
   "quote": "it can in the first attempt, create very sophisticated things that look really good."
  },
  {
   "holder": "Anton Osika",
   "claim": "Even with more capable models like Fable, the binding constraint is the human work of deciding what to build and plugging in the right data — that is not changing as fast as raw model capability.",
   "stance": "opinion",
   "conviction": 0.8,
   "subjects": [
    "Fable",
    "vibe coding"
   ],
   "timestamp": "48:01",
   "quote": "on figuring out what to build, figuring out what are the right strategic directions or experiments you should run to improve the outcomes for your business, that's not changing as fast."
  },
  {
   "holder": "Jason Calacanis",
   "claim": "Paying large token overages is rational because you are still paying only 0.1%-1% of what the equivalent software would have cost a few years ago.",
   "stance": "opinion",
   "conviction": 0.8,
   "subjects": [
    "token economics",
    "Lovable"
   ],
   "timestamp": "44:13",
   "quote": "I'm still paying somewhere between 0.1% and 1% of what I would have paid three years ago. Who cares?"
  }
 ],
 "predictions": [
  {
   "holder": "Pat Gelsinger",
   "prediction": "The AI buildout will be a couple-of-decades-long expansion, not a couple of years.",
   "timeframe": "a couple of decades",
   "subjects": [
    "AI buildout"
   ],
   "timestamp": "18:41",
   "quote": "I am an optimist that we're in a couple of decade build out."
  },
  {
   "holder": "Pat Gelsinger",
   "prediction": "The cost and energy per token will drop by roughly five orders of magnitude, making AI about 10,000x better and vastly more accessible.",
   "timeframe": "unspecified",
   "subjects": [
    "token economics",
    "AI buildout"
   ],
   "timestamp": "18:41",
   "quote": "We want to drop by five orders of magnitude, the cost per token, the energy per token"
  },
  {
   "holder": "Pat Gelsinger",
   "prediction": "Quantum computing will deliver meaningful results across multiple industries before 2030.",
   "timeframe": "before 2030",
   "subjects": [
    "quantum computing"
   ],
   "timestamp": "23:47",
   "quote": "my prediction is meaningful results before 2030",
   "due": "2030-01-01"
  },
  {
   "holder": "Pat Gelsinger",
   "prediction": "Quantum computers will break encryption ('Q-day') roughly in the 2028-2033 window.",
   "timeframe": "2028-2033",
   "subjects": [
    "quantum computing"
   ],
   "timestamp": "23:15",
   "quote": "it's probably going to be 2028, 2032, 2033, when we solve things like encryption, where you'll have the fundamental Q day kind of implications.",
   "due": "2033-01-01"
  },
  {
   "holder": "Anton Osika",
   "prediction": "Within about a year, Lovable will offer an 'AI co-founder' that works while you sleep and returns strategic directions and optimizations each morning.",
   "timeframe": "within a year",
   "subjects": [
    "Lovable"
   ],
   "timestamp": "35:28",
   "quote": "give them access to a co-founder that works for you even when you're sleeping. And it comes back to you in the morning and says like, here are some strategic directions you could go",
   "due": "2027-07-15"
  }
 ],
 "facts": [
  {
   "statement": "Pat Gelsinger spent 34 years at Intel (joining at 18) before becoming its CEO — by his account the first technical leader to run Intel in roughly 15 years.",
   "subjects": [
    "Pat Gelsinger",
    "Intel"
   ],
   "value": "34 years at Intel; first technical CEO in ~15 years",
   "timestamp": "2:34",
   "quote": "when I became CEO in 2001, that was the first technical leader in essentially 15 years",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "In the five to six years before Gelsinger returned, Intel returned about $100 billion to shareholders via dividends and stock buybacks.",
   "subjects": [
    "Intel"
   ],
   "value": "~$100B to shareholders",
   "timestamp": "3:01",
   "quote": "in the five years, five, six years before I came back, you know, Intel gave $100 billion to shareholders.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "Intel had not built a new factory in a decade by the time Gelsinger returned.",
   "subjects": [
    "Intel"
   ],
   "value": "no new fab in ~10 years",
   "timestamp": "3:53",
   "quote": "it hadn't built a new factory in a decade when I got there.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "Apple acquired a small chip company (PA Semi, rendered 'P Semi') to begin building internal silicon competence before launching Apple Silicon.",
   "subjects": [
    "Apple",
    "Apple Silicon"
   ],
   "timestamp": "5:33",
   "quote": "P Semi, you know, they acquired some small company, started to build some competency.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "When Gelsinger came back to Intel, TSMC was producing 5x the wafers of Intel, now more like 7x.",
   "subjects": [
    "TSMC",
    "Intel"
   ],
   "value": "5x -> ~7x TSMC:Intel wafer output",
   "timestamp": "11:32",
   "quote": "When I came back to Intel in 2001, TSMC was producing 5X the wafers of Intel.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "A leading-edge semiconductor fab costs roughly $20-30 billion to build.",
   "subjects": [
    "semiconductors",
    "TSMC"
   ],
   "value": "$20-30B per fab",
   "timestamp": "11:32",
   "quote": "these factories are so expensive, 20 billion, 30 billion.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "When the CHIPS Act started, the US built about 12% of leading-edge chips; today that figure is more like 18%.",
   "subjects": [
    "CHIPS Act",
    "United States",
    "semiconductors"
   ],
   "value": "US leading-edge share ~12% -> ~18%",
   "timestamp": "15:13",
   "quote": "the US was building about 12% of leading edge. Today that number is more like 18%.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "Per a Wall Street Journal article, Taiwan holds less than three weeks of energy reserves, so a blockade could brown out the island without a shot being fired.",
   "subjects": [
    "Taiwan"
   ],
   "value": "<3 weeks energy reserves",
   "timestamp": "15:13",
   "quote": "less than three weeks of energy reserves.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "A semiconductor fab that is powered off does not come back online for about 90 days.",
   "subjects": [
    "semiconductors"
   ],
   "value": "~90-day fab restart",
   "timestamp": "15:13",
   "quote": "When you turn off a fab, it doesn't come back on for 90 days",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "China has blockaded the Taiwan Strait seven times over the last four years.",
   "subjects": [
    "China",
    "Taiwan"
   ],
   "value": "7 Taiwan Strait blockades in 4 years",
   "timestamp": "16:34",
   "quote": "I think China has blockaded the Taiwan Straits seven times over the last four years.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "Global energy capacity is expanding 4-5% while US energy capacity grew about 1% per year for roughly a decade.",
   "subjects": [
    "AI buildout",
    "United States"
   ],
   "value": "global energy +4-5%; US +1%/yr for a decade",
   "timestamp": "18:41",
   "quote": "Energy capacity in the world is expanding 4-5 percent. In the US, we had a decade at 1 percent.",
   "source": "Pat Gelsinger"
  },
  {
   "statement": "Lovable sees roughly one million new products built on the platform every week.",
   "subjects": [
    "Lovable"
   ],
   "value": "1M new products/week",
   "timestamp": "25:43",
   "quote": "We're seeing a million new products built every single week on the platform.",
   "source": "Anton Osika"
  },
  {
   "statement": "Lovable's hosted applications receive more than 700 million visits per month, across more than 50 million apps built on the platform to date.",
   "subjects": [
    "Lovable"
   ],
   "value": "700M+ monthly visits; 50M+ apps",
   "timestamp": "25:50",
   "quote": "more than 700 million visits to the applications every month.",
   "source": "Anton Osika"
  },
  {
   "statement": "About 20% of Lovable's users are technical or engineers; four out of five are non-technical.",
   "subjects": [
    "Lovable"
   ],
   "value": "~20% technical / ~80% non-technical",
   "timestamp": "27:59",
   "quote": "About 20% are technical or some type of engineer.",
   "source": "Anton Osika"
  },
  {
   "statement": "Lovable runs security scans after every change and monitors projects in the background, including for free users.",
   "subjects": [
    "Lovable"
   ],
   "timestamp": "27:59",
   "quote": "run security scans after every change, even now in the background, monitoring the projects.",
   "source": "Anton Osika"
  },
  {
   "statement": "Lovable pricing starts at $25/month, with the business plan around $50/month.",
   "subjects": [
    "Lovable"
   ],
   "value": "from $25/mo (business ~$50/mo)",
   "timestamp": "33:35",
   "quote": "It starts at 25",
   "source": "Anton Osika"
  },
  {
   "statement": "Jason's Founder University intranet, which he estimated would have cost ~$500,000 years ago, was built on Lovable in about four hours for under $2,000 all-in.",
   "subjects": [
    "Lovable",
    "Founder University"
   ],
   "value": "~$500K legacy cost -> <$2,000, ~4 hours",
   "timestamp": "33:37",
   "quote": "It was built in four hours by an employee, which if you just put employees at $50, $60, whatever, $70, plus the cost of your software, it got made for less than $2,000 in a year.",
   "source": "Jason Calacanis"
  },
  {
   "statement": "At Nursa, Nenad used Lovable to build Nursa Study and replace more than 10 internal bespoke tools, saving the company more than $1 million per year.",
   "subjects": [
    "Nursa",
    "Lovable",
    "Nursa Study"
   ],
   "value": "replaced 10+ tools; >$1M/yr saved",
   "timestamp": "37:46",
   "quote": "In their case, they're saving more than a million dollars per year.",
   "source": "Anton Osika"
  },
  {
   "statement": "Lovable reached $500 million in revenue in May, after 20 months in market, having previously grown in roughly $100M increments.",
   "subjects": [
    "Lovable"
   ],
   "value": "$500M revenue (May); 20 months in market",
   "timestamp": "40:10",
   "quote": "We reached 500 in May.",
   "source": "Anton Osika"
  },
  {
   "statement": "Lovable has a strong research team in Stockholm working on post-training.",
   "subjects": [
    "Lovable",
    "Stockholm",
    "post-training"
   ],
   "timestamp": "41:35",
   "quote": "We have a really, really strong research team up in Stockholm who is working on what's called post-training.",
   "source": "Anton Osika"
  },
  {
   "statement": "Roughly 60% of customers on Lovable's lowest subscription tier top up (exceed) their usage caps.",
   "subjects": [
    "Lovable"
   ],
   "value": "~60% of lowest-tier customers top up",
   "timestamp": "44:02",
   "quote": "Something like 60% of our customers, I think...",
   "source": "Anton Osika"
  },
  {
   "statement": "Anton Osika previously worked at CERN, where he encountered the 'co-opetition' model of isolated teams on the same particle accelerator not sharing results until publication.",
   "subjects": [
    "Anton Osika",
    "CERN",
    "co-opetition"
   ],
   "timestamp": "45:54",
   "quote": "for a while I worked at a place called CERN, where they do particle physics.",
   "source": "Anton Osika"
  }
 ],
 "notable_quotes": [
  {
   "speaker": "Pat Gelsinger",
   "timestamp": "3:01",
   "quote": "when you're making these hardcore technical decisions that affect billions of dollars, you don't do that through a spreadsheet.",
   "why": "Crystallizes Gelsinger's entire Intel post-mortem — technology businesses need technologists, not finance leaders, making the calls."
  },
  {
   "speaker": "Pat Gelsinger",
   "timestamp": "15:13",
   "quote": "The economic impact of a brownout of Taiwan is greater than the Great Depression in the world.",
   "why": "The starkest framing of the Taiwan semiconductor supply-chain risk — a blockade needing no shots fired."
  },
  {
   "speaker": "Pat Gelsinger",
   "timestamp": "20:38",
   "quote": "There has not been a time in human history where it's been better to be a technologist than the one we're in right now.",
   "why": "The optimistic capstone of Gelsinger's AI-buildout thesis, set against his bubble caution."
  },
  {
   "speaker": "Pat Gelsinger",
   "timestamp": "22:48",
   "quote": "This decade.",
   "why": "His blunt, falsifiable answer to when quantum computing will finally do something meaningful."
  },
  {
   "speaker": "Anton Osika",
   "timestamp": "30:09",
   "quote": "many engineers, they don't look at the code, they don't write code anymore, and that means that you don't need to be an engineer to create software.",
   "why": "The thesis of the vibe-coding half — software creation decoupled from writing code."
  },
  {
   "speaker": "Jason Calacanis",
   "timestamp": "44:13",
   "quote": "I'm still paying somewhere between 0.1% and 1% of what I would have paid three years ago. Who cares?",
   "why": "The economic-leverage argument for why token overages are worth paying."
  }
 ],
 "extracted_by": "agent",
 "prompt_version": "episode-extract@3",
 "_file": "all-in-2026-07-15-former-intel-ceo-on-what-went-wrong-what-s-next-lovable-ceo-.json"
}