TBPN Podcast – February 24, 2026

Episode Theme:
A real-time pulse check on the tech and finance world as markets reel from the “Citrini Memo,” with industry-shaking conversations on AI’s impact on SaaS, tech research paradigms, consumer product launches (Kim K’s energy drink), regulatory scrutiny (Jane Street’s lawsuit), and a star-studded parade of founders and investors weighing in. Guests include the Collison brothers (Stripe), Bill Gurley, Ivan Zhao (Notion), Scott Wu (Cognition), James Cadwallader (Profound), Reiner Pope (MatX), Stefano Ermon (Inception Labs), Runa Kvist (Artificial Intelligence Underwriting), and more.

Key Topics & Discussions

The Citrini Memo Aftershock: Markets and Narratives

Timestamps: 00:00–24:00, 36:00–52:00

• Live Market Reaction:
  • “Today we are surviving the Citrini apocalypse. Live to fight another day. A lot of chaos in the markets… But it didn’t slap the markets for it.” – Host A [00:00]
• Rise of Independent Tech Analysts:
  • Conversation on how Substack, X (Twitter), and independent analysts (e.g., Ben Thompson, SemiAnalysis) increasingly shape investor sentiment.
  • “We are the sell-side research now. Basically, we as in substack, but X and substack, independent researchers and analysts are really moving the markets.” – [03:30]
• Meta-Commentary from Derek Thompson:
  • “The conversation about AI is a marketplace of competing science fiction narratives… When you drill down underneath the bluster, the boosterism, the fear, the anxiety, what’s there at the bottom is genuine uncertainty, a vacuum into which storytelling is flooding.” [14:50]
• AI Doomerism vs. Real Tech Progress:
  • Hosts challenge the pessimistic scenario: “If there’s a 5% chance that everything’s cooked, the market should probably sell off by a couple percent. And you know, the market didn’t even really sell off.” – [09:00]
• Timeline of AI Evolution:
  • Debate around “continuum” of AI and that most tech predictions eventually come true—just usually on much longer time scales (20 years vs 2 years).
  • “I’m impatient about it. I want it to go faster, I want the acceleration, I want the progress. I think progress is good.” – [18:55]

The SaaS-copocalypse & Open Source Pressures

Timestamps: 24:00–48:00

• Contradictory Valuations:
  • Private AI-native SaaS startups booming and attractive to investors even as legacy SaaS stocks drop in public markets.
  • “Private companies getting lots and lots of funding... that if they were public would have traded down 20% over the last week.” – [25:30]
• Open Source/“Vibe Coding” vs Managed SaaS:
  • Theoretical pressure from open source and agent-based coding: “Maybe like, if you can just prompt it and it feels like emailing your SaaS provider to reconfigure things, that is a real pressure.” – [29:30]
• Why Open Source Hasn’t Fully Disrupted SaaS:
  • “It’s just annoying to maintain. So no one ever does it and you’re kind of just paying whoever to maintain it, right?” – [31:00]
• Agents and the Future of Software:
  • “Everyone will have personalized software... but also, the lab providing the software, the inference.” – [33:00]
  • Competing narratives: will individuals run their own agentic software or will value accrue to the platforms?

DoorDash as a Case Study: Defending Tech Value

Timestamps: 48:00–52:00

• Ben Thompson’s Take:
  • Pushes back on the notion that DoorDash will be made obsolete by AI: “DoorDash is the ‘I’m hungry’ button on your phone… There’s no awareness that DoorDash provided a massive consumer benefit from scratch…” [49:00]
• Role of Human Value:
  • Case of real estate agents—why some tech-enabled disruptions haven’t materialized: “The truth is the Internet already obsoleted real estate agents in terms of information flow… yet humans will be resourceful in giving themselves jobs to do.” [52:00]

Kim Kardashian in CPG, Meme Culture & LLM Training

Timestamps: 61:00–65:00

• Kim Kardashian’s ‘Update’ Energy Drink:
  • Launch analyzed, “If Kim Kardashian is coming for your consumer product brand, I feel like you’re in trouble. She’s almost a lawyer.” [62:30]
• Info Work Deflation, AI, and Competitive Moats:
  • “Information work itself has a real premium in pricing power… that one way trade can go backwards really hard all at once…” – Doug, Fabricated Knowledge [64:40]
• Jokes and Dystopian Scenarios:
  • “Telling my kids that if they don’t clean their rooms, Citrini will come for them. Dangerous stuff.” – [67:00]

Guest Segments & Notable Insights

John & Patrick Collison (Stripe)

Timestamps: 74:00–101:00

• Stripe News Drop:
  • Announcing a tender offer for employees and release of their annual letter (34% YoY growth).
• On Real Economy vs. Tech Market Sentiment:
  • “If you look at the actual real economy time series… over the last two years, things really seem in good shape.” – Patrick Collison [77:20]
• Signs of the AI Boom in Startup Cohorts:
  • “There has been a phase transition in 2025… more [businesses starting] and on a per-business basis they are on average doing better… 2026 may plausibly be an acceleration even over that significant leap in 2025.” – Patrick Collison [82:00]
• Agentic Commerce & Payment Innovation:
  • Stripe positioning for an explosion in “agentic commerce”—agents using both legacy rails and stablecoins.
  • “We think you’re going to need blockchains and better blockchains, honestly… you’re going to need really high throughput ones for the agents.” [84:50]
• Product Strategy Wisdom:
  • “You can’t get too MBA brain about new products… reason in product specifics.” – John Collison [92:00]
• Notable Moment:
  • Stripe Press sold their 1.1 millionth book. “Books… they’re very AGI-proof.” [98:30]

Bill Gurley (Venture Investor – “Running Down a Dream”)

Timestamps: 102:00–124:00

• Book Launch and Career Reflections:
  • Gurley’s new book is directed at helping people find meaningful careers, not just VC insiders.
  • “One of the problems… is that our common college pathway has actually become more restrictive… not a lot of exploration or search for creativity or obsession.” [111:45]
• AI and Public Narrative:
  • “I don’t remember one… being thrown at the public consciousness this fast.” [105:00]
• Hyper-Financialization & Bubbles:
  • “If you wanted AI exposure, you’re pretty good just owning the index. Nvidia is such a large part…” [118:00]
• Venture & Staying Private Longer:
  • “They hijack the growth years of these early IPO companies. Amazon went public below a billion in market cap… It’s hard to fathom that today.” [121:00]
• On China & Open Source:
  • “The biggest threat to the US… is the Chinese open source models. That’s going to be a really incredible primordial soup for innovation to evolve.” [124:00]

Ivan Zhao (Notion) – Notion Custom Agent Launch

Timestamps: 147:00–162:00

• Major Product Ship – Custom Agents:
  • “It does real work for you in the background, very easy to set up, host in the cloud… you don’t need a Mac Mini.” [148:00]
• Notion as Agent Orchestration Layer:
  • Use case: Ramp (startup) consolidated half a dozen collaboration tools into Notion – “Now they have one place to do their core collaboration work, one place to deploy AI.” [150:00]
• Multiplayer, Enterprise-Grade AI Automation:
  • “Cron job is a pretty good word… a lot of knowledge work is just cron job. Update, push paper back and forth, from this person to the other.” [153:00]
• AI & “No Code” Philosophy:
  • “We always care about building beautiful and powerful tools… AI is that today. How do we make sure that all company can benefit from this?” [160:00]
• Memorable Quote:
  • “Let AI do the night shift so you can do the day shift.” [154:00]

Profound (James Cadwallader) – Marketing Agents & GEO

Timestamps: 163:00–173:00

• $96M Series C, Unicorn Status
• Marketing Agent Platform:
  • “Harvey lawyers have Harvey, engineers have Cursor, marketers have Profound.”
• On Fixing LLM Hallucinations About Brands:
  • “...the models were incorrectly spitting out that there was no FDIC insurance... we identified it was coming from a blog, Reddit posts... We reached out and corrected all three.” [169:00]
• GEO (Generative Engine Optimization) Discussion:
  • “GEO sucks as an acronym… ‘Answer engine optimization’ feels more fitting…”

Cognition (Scott Wu) – Devin/AI Agents

Timestamps: 173:40–186:00

• Enterprise AI Adoption Doubled in 6 Weeks:
  • “Mass takeoff of agents… what you really need is the full background cloud agent.” [174:00]
• Product Improvements:
  • “Automated testing, having Devin run your web app... send changes, screencaps… VM startup way faster…”
• AI Progress:
  • “I think things are continuing on the exponential curve… but now the form factor looks different as you continue. You’re actually solving different problems.” [183:00]
• On SaaS Apocalypse:
  • “Software is one of the most deflationary things ever... the same products that used to cost much more have gotten much, much cheaper. Has this been terrible for software companies? It’s been pretty good.” [178:30]
• Memorable Q:
  • “Will AGI be here when I get to have a pizza that tastes amazing and also is fully nutritious?” [185:20]

Stefano Ermon (Inception Labs) – Diffusion LLMs

Timestamps: 138:00–147:00

• Diffusion Models for LLMs:
  • “Start with a rough guess of the answer and refine it… The difference is that the neural network is able to modify many tokens at the same time, so it’s much more efficient.” [139:00]
• AI Voice and Code Use Cases:
  • “We can get to over a thousand tokens per second on Nvidia GPUs. Latency-sensitive apps, code suggestions, voice agents, retrieval search…” [142:00]

Rune Kvist (AI Underwriting) – Insuring AI Agents

Timestamps: 186:00–194:00

• First Ever Insurance Policy for AI Agents (w/ 11 Labs):
• On Managing New AI Risks:
  • “Regardless of whether anyone buys an insurance product, someone is always underwriting it…”
  • “Eventually we think some of the risks that look a little more like private nuclear energy will have to be covered by insurance.” [188:00]
• Regulatory Models:
  • Reference to the Price-Anderson Act for nuclear as precedent for government backstop.

Reiner Pope (MatX) – Custom Silicon for LLMs

Timestamps: 194:00–203:00

• $500M Fundraise – Jane Street & Situational Awareness Lead
• On Building Fit-for-LLM Chips:
  • “If you want to make the best chips for LLMs—design for large matrices, low precision, low latency—you need to start from scratch.” [195:30]
• Nvidia Cuda Lock-in and Opportunity:
  • “CUDA is simultaneously the biggest advantage and also a constraint… It means the next generation GPU has to look just like the GPU from 10 years ago.” [197:20]
• Investor Powerhouse:
  • “It’s the hardest lineup ever heard—Jane Street, Situational Awareness, Andrej Karpathy, the Collisons...” [200:00]

Devanche Pandey (Standard Intelligence) – Computer Use Models

Timestamps: 203:00–212:00

• Pre-training on 30fps Video of Computer Use:
  • “We have this app that people run… It records their screen, logs all key presses and mouse movements.” [204:00]
• Short-Term Use Cases:
  • “Mechanical engineers doing CAD, can have their next minute of manual work done. We showed gear extrusion in CAD—LLMs are really bad at CAD.” [208:30]
• On Text vs. Video as Interface:
  • “There are just a lot of things much more native when done visually… There’s a reason humans don’t interact with the computer purely through text.” [210:30]

Running Timeline Highlights

• Jane Street Sued over Terra/Luna:
  • Allegations of insider trading—group chat evidence, rapid withdrawals mirroring Terraform Labs, questions over how public blockchains may still enable information asymmetry or action. [124:40]
• Meta x AMD:
  • Meta will integrate AMD Instinct GPUs at huge data center scale for AI model acceleration. [157:00]
• Deepseek Blackwell Ban Circumvention:
  • Report that Deepseek, a leading Chinese lab, trained its model using banned Nvidia Blackwell GPUs. Panel is skeptical that Chinese labs will ever truly leapfrog the US. [165:00]
• Miscellaneous Tech News:
  • Supreme-branded Shure MV7 podcast mic announced, engaging podcaster/creator communities.
  • McDonald’s launches “Big Arch” burger—US CPG novelty.
  • Capybara Simulator goes viral as another meme game in genAI era (“Conor McGregor says: take my money!”) [158:00]
  • HubSpot acquires Starter Story—evidence of SEO-to-video platform transition. [211:30]
  • Stripe rumors to acquire PayPal surface live on air. [180:00]
  • Arena Magazine moves to Substack—indie tech media adapt to new distribution.

Quotes & Memorable Moments

• On AI Uncertainty:

  “Nobody knows anything except for us is about as close to the reality here as three words are going to get you. Nobody, nobody. What’s—nobody knows what’s going to happen this year or next year or the year after that.”
  — Derek Thompson [16:00]

• Stripe on AI & Real-World Data:

  “When we look at the cohorts… the businesses that signed up in 2025… both more of them and the median business is performing better. There has been a phase transition.”
  — Patrick Collison [82:00]

• Notion’s Evolution:

  “Notion started as a computing tool… The power of computing is in the hands of you, the programmers. How do we open up to more people… Now language models can use Notion’s building blocks to do work for them.”
  — Ivan Zhao [157:00]

• Bill Gurley, on Agency and Careers:

  “Once they latch on… we have examples in the book where that doesn’t happen till 40. Sometimes it’s at 30... I didn’t become a venture capitalist until I was 30, and that was clearly my dream job.” [112:00]

  “There’s never, ever in the history of the world been a better time to self-learn. It’s like magic.” [116:00]

• Devanche Pandey on Training Data:

  “You have a bunch of data of humans doing things… a lot of times humans make mistakes and then have to correct those. And you don’t get that with text, because you don’t get to see the process of messing up and fixing it.” [211:00]

• On LLM Progress:

  “We stopped typing code at some point… The code we check into GitHub, how much of it was typed by a human? Almost none.”
  — Scott Wu [183:00]

Important Timestamps

• [00:00–36:00] – Citrini Memo & Sell-side Shockwaves
• [36:00–74:00] – Legacy SaaS, Agents, and Agents-as-Software
• [74:00–101:00] – Stripe: AI Impact, Cohort Data, Agentic Commerce
• [102:00–124:00] – Bill Gurley: Finding Life’s Work, AI Wave, VC Industry
• [147:00–162:00] – Notion: AI Agents for Knowledge Work
• [163:00–173:00] – Profound: Agents & Marketing
• [173:40–186:00] – Cognition: Speed, AI Agent Adoption
• [194:00–203:00] – MatX: Custom Silicon, $500M Raise
• [203:00–212:00] – Standard Intelligence: Video ML, Computer Use Models

Closing & Meta Moments

Culture:
The episode regularly lampoons hype cycles (“Citrini apocalypse,” “royal flush” of investor lineups), tech Twitter memes, “vibe coding,” and sponsors that make up the podcast’s satirical, fast-paced, nerdy, self-aware tone.

For New Listeners:
This episode is a window into how top tech founders and investors process disruption, meme culture, and real economic data—capturing the blend of existential uncertainty, historic shifts, and rampant optimism that defines today’s Silicon Valley moment.

“Have the best evening of your entire life. We love you. Goodbye.” [212:00]