Nov 25, 2025 · 26m · tbpn
Reverse Engineering 200 AI Startups, Nucleus Genomics Controversy, Drone Hunting | Diet TBPN
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
John Coogan and his co-hosts analyze major developments in the tech ecosystem, including Gemini 3's battle against ChatGPT, OpenAI's hardware ambitions, deceptive AI startup wrappers, controversial biotech marketing, and autonomous drone delivery regulations.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 74.6% of the talking time here. How this is scored →
The hosts as informed peer 5.3 Guest teaching 1.4 Guest disagreement 0.8 The hosts pushing back 0.6
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 4:42 Tyler challenges John's definition of multimodality
Tyler directly questions John's assertion that Gemini is not genuinely multimodal, arguing that all modalities are unified into tokens at the architecture layer.
Hardest push from the hosts ▶ 4:44 John insists explicit mode buttons reveal underlying product limitsJohn rejects the technical architecture defense, pushing back that requiring manual UI selectors proves the user experience is not truly seamless or multimodal.
Biggest teaching moment ▶ 4:49 Tyler clarifies underlying token architectureTyler educates the discussion by distinguishing underlying transformer token processing from consumer frontend implementation quirks.
The host holds their own ▶ 14:20 John explains exponential backoff rate-limit fingerprintingJohn demonstrates deep technical literacy by outlining how unique exponential retry curves and network traces uncover startups quietly wrapping OpenAI APIs.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Evaluating Gemini 3 as a Daily Driver Against ChatGPT | 6 | 1 | 0 | 0 | John opens with a detailed breakdown of his daily driver switch to Gemini 3, analyzing model latency, routing nuances, pricing tiers, and response sizing compared to GPT-5 Pro. The co-hosts engage in a collaborative and aligned technical discussion. | |
| Examining Consumer App Friction and Multimodal Limitations | 5 | 3 | 3 | 3 | John criticizes Gemini's UI for lacking seamless multimodal transitions, prompting Tyler to push back and clarify that token-level multimodal architecture is distinct from UI buttons. John defends his user-experience framing by asserting that separate mode buttons expose underlying system limitations. | |
| AI Monopolies, Consumer Lock-in, and Alphabet's Financial Resurgence | 6 | 1 | 0 | 0 | The host analyzes the fragility of ChatGPT's consumer lock-in, highlighting that memory and custom instructions did not prevent his platform switch. The co-hosts review Sergey Brin's media appearance and Alphabet's two-trillion-dollar market cap rebound. | |
| OpenAI Poaches Apple Engineers for Secretive Hardware Venture | 5 | 2 | 1 | 1 | John contextualizes Mark Gurman's reporting on OpenAI poaching roughly 40 Apple hardware engineers for Jony Ive's startup project. The co-host looks up Apple's overall hardware headcount to frame the poaching numbers in proper perspective. | |
| Sam Altman Reflects on the OpenAI Boardroom Turnaround | 7 | 1 | 0 | 0 | John delivers an in-depth technical explanation of how an investigation reverse-engineered 200 AI startups using Chrome network traces and exponential backoff API fingerprinting. The co-hosts agree on the prevalence of thin LLM wrapper startups in venture portfolios. | |
| Nucleus Genomics Sparks Controversy with IVF Subway Campaign | 4 | 3 | 2 | 1 | The host breaks down Nucleus Genomics' viral subway campaign as deliberate rage-bait marketing. The co-host adds crucial context by noting unsubstantiated competitor accusations and ethical problems surrounding potentially fictitious customer reviews. | |
| Hyper-Realistic AI Images and Satirical Online Personas | 5 | 0 | 0 | 0 | John details how Google's SynthID watermarking and contrast inspection can identify AI-generated imagery. The co-hosts share a lighthearted discussion reviewing satirical AI profiles and viral deepfakes. | |
| Bill Gurley Honored with Texas Distinguished Alumnus Award | 4 | 0 | 0 | 0 | The episode concludes with banter about Bill Gurley's alumni award and Zipline's autonomous food delivery drones. John jokes about drone-hunting burritos while noting the severe federal FAA legal consequences. |
Statements from this episode (11)
Coogan pays $250 monthly for Google Veo 3 due to lacking competition
“VO three was the moment when I was like, okay, they got something that nobody else has. I got a fork over two 50 a month.”
Coogan switched to daily driving the Gemini iOS app for knowledge retrieval
“I switched to daily driving Gemini on iOS as the main app that I go to for all the different knowledge retrieval requests.”
Gemini 3 thinking queries match GPT-5 Pro reliability with less reasoning
“I feel like the level of confidence in the brand makes me feel that a Gemini three thinking query that does maybe less reasoning Then a GPT five pro query will be at the same level of reliability.”
Google Gemini's fragmented UI forces manual mode selection before prompting
“What I'm saying is that we are still in the pre like selected dropdown UI functionality of Gemini, because I, I'm prompted to pick what I want to do. Do you want to do image, video, deep research, text before you go into the flow instead of just saying, I'm ha…”
Real-time voice AI is not the preferred user interaction method
“I've used it. I've used it a bunch. I've used all of them, but they're just not the preferred way of interacting.”
The consumer AI market will shift from winner-take-all to a duopoly
“Maybe, maybe there will be a little bit more of a duopoly. There, it won't be such a winner-take-all market even though it has been historically, it's, it has been up to this date.”
Google added $2 trillion in market cap since the Gemini image controversy
“Google has now added two trillion dollars to its market cap over the past 20 months since the boob shirt guy asked Sergey Brin about woke Gemini images while having a foot-long Subway cold-cut trio for lunch.”
Sergey Brin admits Google's Gemini image issues stemmed from inadequate testing
“You know, we definitely messed up on the image generation and I think it was mostly due to just like not thorough testing.”
OpenAI's upcoming hardware device is not yet a serious threat to Apple
“Now they're clearly going to survive there. It's not a serious threat, at least not yet. Maybe if this device is incredible, but right now Apple looks pretty strong from an AI perspective.”
OpenAI's unique API rate-limiting curves expose companies using it under the hood
“The shape of that curve and the specific timings are unique to open AI.”
Google Gemini uses SynthID detection to identify origins of AI-generated images
“If you put the photo into Gemini Gemini will tell you, if you say, is this AI, Gemini will tell you, yes, according to the SynthID water molecule detection tool, this image was generated in whole or in part with Google AI.”