TBPN Podcast Summary

Episode Title: AI Adoption Trending Down for Large Companies, OpenAI Backs AI Animated Feature Film, China's Smart Glasses Domination
Hosts: John Coogan & Jordi Hays
Date: September 8, 2025
Notable Guests: Avi Schiffmann (Friend), Scott Wu (Cognition), Ara Kharazian (Ramp), Sunny Madra (Groq), Zach Lloyd (Warp), Alex Cohen (Patient), Mati Staniszewski (11 Labs), Daniel Kao (Truck Smarter)

Overview

In this jam-packed episode of TBPN, John and Jordi are joined by a slew of leading tech founders and analysts to dissect trending technology stories: the purported decline of AI adoption in large companies per the U.S. Census, OpenAI’s bold venture into feature-length AI-animated movies, and the competitive pressure China is placing on Western companies in the smart glasses race. The discussion includes live reactions to viral data, detailed breakdowns of the startup fundraising environment, and in-depth founder interviews, all delivered with the pod’s trademark mix of authenticity, skepticism, and meme-worthy banter.

Main Themes and Segment Highlights

1. AI Adoption in Large Companies: Trend or Mirage?

(00:12–16:43, 134:35–148:51)

• U.S. Census Data Claims AI Adoption is Sliding in Large Firms
  • Viral chart from U.S. Census’ bi-weekly business survey: AI usage among companies with 250+ employees fell from 14% to 12%.
  • John: “The chart is going viral because it's dropping off and it looks bearish for AI and thus for tech America, for humanity broadly. Everyone is saying, please consult the Gartner hype cycle graph.” [06:10]
  • Jordi: “Small firms haven't declined even by this metric...firms with one to four employees are still increasing in AI adoption.” [07:07]
  • Discussion questions Census’ definition of “AI use,” which includes things as broad as image processing and predictive analytics, and why data seems at odds with high consumer LLM adoption rates.
• Ramp’s Counter-Data: Business Spending on AI is Rising
  • Ara Kharazian (Ramp): “When we look at ramp spend data, AI adoption is up. AI adoption is now about 45% of businesses in the U.S...Even in restaurants, we see 20% adoption of AI.” [139:07]
  • Ara critiques: “The Census measures [AI] by asking: do you use AI to produce goods or services? ...That's econ speak, not how actual operators think about their work.” [135:06]

Notable Quotes

• John: “It makes no sense that 10% of companies would fit this definition...If your business processes receipts with image recognition, you're technically an AI user—but you don't see yourself that way.” [12:02]
• Ara Kharazian: “Sample sizes for these surveys are actually pretty small...the methodology is noisy at best.” [138:03]

2. OpenAI’s Feature-Length AI Film: “Critters”

(17:42–42:27, 181:20–183:10)

• OpenAI Announces Animated Movie, 'Critters'
  • $30M budget, will be shown in theaters, aiming for Cannes debut.
  • AI is used for much of the animation, but human actors to provide voices and artists for sketches.
  • John: “OpenAI's betting that if Critters is successful, it will show that AI can deliver strong content...and accelerate Hollywood's adoption of the technology.” [41:33]
• Discussion: AI’s Hollywood Moment, Or Gimmick?
  • Tension around using real voice actors highlights the "uncanny valley" of synthetic voices—even as companies like 11 Labs lead in the space.
  • Hosts compare to the Sonic the Hedgehog redesign, suggesting audience pushback could steer AI-driven media.

Notable Quotes

• Jordi: “This feels like something they can allocate some capital, prove a point...They're spending $30 million with a cool creator and sending it over there.” [20:26]
• Mati Staniszewski (11 Labs): “The true voice experience takes an artist...Across Hollywood, aligning likeness and compensation with AI-generated voices is still an open question.” [179:54, 181:20]

3. The “996” Work Phenomenon, Startup Trends, and The Great Lock-In

(141:00–148:51, 49:24–50:25, 141:40–147:52)

• 996 ("9am–9pm, 6 Days a Week") Makes a Comeback in SF Startups
  • Ramp’s internal data: Sharp pickup in business meal expense activity on Saturdays among SF startups—a concrete sign the meme is real, at least locally.
  • Ara: “This wasn't happening last year. Perfectly coincides with when people started talking about the 996 cultural movement.” [142:30-142:43]
  • This is not matched in NYC, Austin, or Miami.
  • Hosts reflect on the startup attitude shift post-layoffs and the flood of young founders skipping college due to accessible AI tooling.
• Startup Fundraising Frenzy (“The Great Lock-in”)
  • “There are so many serious fundraises...tons of great companies. We're officially back!” [15:01]
  • Guests Avi Schiffmann (Friend), Scott Wu (Cognition), Zach Lloyd (Warp), and others break news of new fundraising rounds and discuss competitive startup strategy.

4. AI Tools: Genuine Productivity or Vibe Coding Chaos?

(03:21–05:47, 115:50–119:05, 153:09–155:38)

• Mixed Findings on Developer Productivity with AI Code Tools
  • Some studies claim AI coding tools can double code output but also increase security vulnerabilities by 10x.
  • Scott Wu (Cognition): “For a lot of these tasks...We see speedups of around 8 to 15x. Devin (our agent) is not necessarily optimized for ideation, but for engineering toil.” [116:26, 114:19]
  • Zach Lloyd (Warp): “They [agents] kind of produce stuff that's not shippable a lot of the time...It's really a mess to vibe code in production right now.” [151:50, 153:09]
  • Discussion of the emerging job category: “Vibe code cleanup specialist.”
  • Both founders say disciplined workflows, code review, and new tooling will be critical to avoid “messy agent spaghetti.”

Notable Quotes

• Scott Wu: “Security is a problem all orgs have already had to deal with...You want similar processes in place for your AI agents.” [118:08]
• Jordi: “There's always a bull market somewhere—with vibe coding cleanup, the bull market is now.” [04:01]

5. China’s Smart Glasses Surge and Meta’s Competitive Dilemma

(70:12–76:45)

• Smart Glasses Pricing Pressure
  • Meta’s upcoming “Hypernova” smart glasses, once rumored to be $1,400+, will now launch at $800, likely in response to $280 glasses from Alibaba’s Quark Vision.
  • Breakdown of "smart glasses" generations, with Gen 2 (low-res displays) set to become the true inflection point for mass adoption.
  • Consensus among panel: AR/VR displays still face physics constraints; but the rapid price drop driven by Chinese players will reshape the market.
• Takeaway:
  • John: “If China is putting products out this cheaply, Meta can’t justify a $1,400 price...Smart glasses may be the next smartphone class, but display fidelity and use cases will decide the winners.” [71:46]

6. AI Agents as SaaS: Reframing the Technology

(16:28–16:50, 166:11–167:25)

• Reframing ‘AI’ as SaaS
  • Many will use AI and stop considering themselves “AI users”—as ML becomes just another SaaS component.
  • John: “AI is going to seep into every crack of small business operations...but the operator won’t be proudly saying ‘I’m all in on AI’ forever.” [16:50]
• Vertical SaaS Will Win in Specialized Categories
  • Alex Cohen (Patient): “Healthcare AI must be custom, deep vertical SaaS. Clinics don’t care it’s AI—they care if it works and fits their workflow. That's product-market fit.” [167:25]
  • “Sell the solution, not the technology.”

7. Other Noteworthy Segments & Founders

(21:34–28:44, 149:26–156:53)

• Avi Schiffmann (Friend): Wired Review, Voice, and AI Personality
  • Defends his “opinionated” AI friend device after a scathing review; sees solid user engagement and hints at creative monetization like “life insurance” for AI companions.
  • Avi: “They're programmed to make you more confident and more agentic, which the world needs more of...If you have a bad time with Friend, it might not be the product, it might just be you.” [23:32]
• Zach Lloyd (Warp): Hat Launch and Agents in Development Tools
  • Discusses “Code Country” launch (“wrangle your agents”) and the challenge of producing robust, debuggable agent-generated code.
• Mati Staniszewski (11 Labs): Growth and Secondary at $6.6B
  • Announcing a $100M secondary at $6.6B—a move to accommodate early employees and “align for the long haul”.
  • Mati: “Jensen Huang told me: ‘The true voice experience takes an artist. You’re all artists at Eleven Labs.’ Voice will not commoditize as fast as people think.” [179:54]
• Daniel Kao (Truck Smarter): Load Boards and SaaS in Freight
  • Raised $16M, aiming to disrupt paid trucking marketplaces with a free “Craigslist for loads” and financial products for independent truckers.

Notable Quotes & Timestamps

• “I don't know if I'm on the left side or right side of the Gartner hype cycle anymore. 2024 felt like the trough of disillusionment.”
  — John [05:03]

• “AI adoption is going to become invisible. You won't be an 'AI user' — you'll be a SaaS user.”
  — John [16:28]

• “You want to know who is using AI in business? The ones with access to paid Ramp receipts or OpenAI subscriptions — and it's rising, not falling.”
  — Ara Kharazian [139:07]

• “We’re just scratching the surface of what agents can do. But you can’t ship vibe code straight to production — not yet.”
  — Zach Lloyd (Warp) [153:09]

• “Across Hollywood, aligning likeness and compensation with AI-generated voices is still an open question.”
  — Mati Staniszewski (11 Labs) [179:54]

• “Healthcare ops is not a generic chatbot problem — we have to deeply integrate with messy clinical realities, and no big platform is going to do that for you.”
  — Alex Cohen (Patient) [159:15]

Timestamps of Key Segments

• AI adoption decline chart, Census v. Ramp data: [00:12–16:43, 134:35–148:51]
• OpenAI feature film, uncanny valley discussion: [17:42–42:27, 181:20–183:10]
• Vibe code, agent workflows, Warp & Cognition on coding: [03:21–05:47, 115:50–119:05, 153:09–155:38]
• 996 trend, startup fundraising “great lock-in”: [141:00–148:51, 49:24–50:25]
• Smart glasses industry, Meta/China price war: [70:12–76:45]
• Founders: Friend, Patient, 11 Labs, Truck Smarter: [21:34–28:44, 149:26–156:53]

Tone & Language

The hosts maintain an extremely conversational, tongue-in-cheek, and sometimes meme-driven style but always loop back to sharp operational insights. Founder guests are direct and forthright. There's a sense of competitive energy, optimism for technology's future, and an undercurrent of skepticism toward media hype/fear cycles.

For Listeners Who Didn't Tune In

• Don’t buy the “AI winter” meme—AI adoption is shifting, not shrinking, and industry definitions don’t match reality.
• OpenAI’s animated movie signals a coming wave of consumer-facing AI creativity, but human touch isn’t going away.
• Many startups are re-finding their “996” hustle—at least in SF—while young founders flood in thanks to AI tools.
• Agentic development and “vibe code” will be huge, but require new guardrails and cleanup as non-specialists create foundational code.
• Chinese hardware price pressure is real; software differentiation, distribution, and focus will be the edge for Western companies.
• Leading founders stress: It’s about solving the problem (not selling “AI”), building vertical SaaS, and adapting the stack to new realities.

For further details, tune into the next episode or revisit these conversations with any of the featured founders on X or TBPN’s platforms.