(Preview) An OpenAI Reminder, Netflix’s Expanding Appetite, Q&A on Remote Work, Taco Bell, and Data Centers in Space
Summary
ChatGPT Image 1.5 narrows the image-generation gap that had been giving users a reason to open Gemini, but Ben Thompson’s more important call is that OpenAI keeps winning at product. Gemini produced his single favorite image, yet ChatGPT won across the set and wrapped existing capabilities in an interface that helps ordinary users discover what AI can do: “You have to lead them down that path.” The same week’s GPT-5.2 update used the same base model with more post-training; Ben called it a good response ahead of a supposedly next-generation model in the new year.
ChatGPT’s contextual continuity turned image generation from a tool awaiting instructions into a product that stimulated more demand. Reusing identical prompts, Ben found that ChatGPT automatically preserved characters and visual style across images, creating a comic-book-like thread that “was pulling ideas out of me”; with Gemini and Nano Banana, he generally started only after already having an idea.
OpenAI’s popularity is simultaneously its moat and a compute-intensive catch-22. A ChatGPT feature may reach 900 million people, and AI retains a marginal cost, so deployment resembles Apple manufacturing Face ID at tens-of-millions scale: invention is insufficient without reproducibility. Ben revived his monetization critique—“I should’ve been even harsher on ads”—arguing that a 2023 launch would have produced a mature ad product and habituated users by now.
Disney licensing characters to Sora may not revive the app, but it reinforces the rising value of universally recognized IP. Ben called the agreement bullish for Disney “even if Sora doesn’t amount to anything”: when AI makes content highly customized and individualized, common cultural “totems” become more valuable because audiences need shared reference points.
The AI competitive map increasingly looks like vertically integrated Google against a coalition, although Google’s economics remain formidable. Disney’s Sora agreement and near-simultaneous cease-and-desist to Google place OpenAI, NVIDIA, and Disney on one side, while Google owns its chips and models and dislikes deals because “deals don’t scale.” Yet Gemini 3 Flash is “extremely cheap” and strong enough that its API economics may matter more than flagship-model rankings.
Ben’s strategic case for Netflix’s Warner Brothers pursuit is built around attention, where YouTube is the central threat, while narrower markets remain relevant antitrust questions. Netflix is working to pull at least 18 podcasts’ video episodes from YouTube through iHeartMedia and Barstool deals beginning in early 2026—another step toward becoming a stand-in for all television and entertainment.
Deep dive
1. OpenAI narrowed Gemini’s clearest consumer advantage
Ben’s practical verdict on ChatGPT Image 1.5 was measured: Gemini generated his single favorite image, but he preferred ChatGPT’s output overall. Both are “amazing”; strategically, OpenAI needed to shore up images because they had become a reason for existing ChatGPT users to leave for Gemini.
Ben’s analogy captured the prior division of labor, with Andrew agreeing: “If you wanna generate a picture of a turkey, go to Gemini. If you want the correct instructions for how to cook it, go to ChatGPT.” Andrew became curious about Nano Banana only after a friend showed him impressive Photoshops made with it, and he still had not tried it when ChatGPT released the feature a day later.
The release’s dedicated image area, example prompts, and suggested projects mattered more than their technical simplicity. Ben’s broader product thesis is that technical people assume capability will sell itself, but users “don’t know what’s possible,” so a product must teach them—an instinct he believes has distinguished ChatGPT over the past 2.5 years.
The release also arrived alongside GPT-5.2, which Ben described as the same base model with more post-training. He called it a good response while waiting for the supposedly real next-generation model in the new year.
2. Context turned image generation into a compounding creative loop
Replaying a long Gemini thread word for word inside ChatGPT produced an unexpected advantage: repeated characters stayed consistent and the visual style remained the same without Ben requesting either. The result felt like illustrating an ongoing comic book rather than generating disconnected assets.
That continuity altered user behavior. With Gemini and Nano Banana, Ben arrived only after conceiving an image; ChatGPT “was pulling ideas out of me,” making him want to extend the story. His product test therefore moves beyond capability and discovery: “Can it inspire you? Can it make you want to use it more?”
Andrew supplied the consumer evidence, spending about an hour generating his children in different scenes. Ben’s competitive conclusion was blunt: This product advantage may not sound like a moat because Google could copy it, “but they don’t. They don’t again, and again, and again.”
3. OpenAI’s consumer scale magnifies compute and monetization mistakes
Ben argued that OpenAI should let Microsoft handle the API and focus on delivering “the most compelling user experience” in the consumer app it has a chance to rule the world with. He regretted softening that critique: “I feel bad that I backed down on my critique there, ’cause I was totally right.”
His second reversal concerned monetization: “I should’ve been even harsher on ads.” Had OpenAI introduced advertising in 2023, he believes the product would now be good and consumers accustomed to it.
Popularity creates a catch-22: it is OpenAI’s greatest advantage and possibly its biggest problem. Ben compared deployment to Apple’s struggle to manufacture Face ID reliably at tens-of-millions scale; because AI carries marginal cost, a ChatGPT feature can reach 900 million people, making rollout “really, really hard,” even when the underlying technology works.
4. Disney’s Sora deal raises the value of shared IP
Ben was unsure whether Sora would become a lasting thing—he noted that it was number 25 in the App Store—while Andrew was skeptical that adding Darth Vader would revive it. Ben nevertheless viewed OpenAI’s ability to create multiple viral moments as a positive signal. Early Sora content centered on Sam Altman because its tech-heavy users all recognized him; the lack of copyright enforcement also made absurd Pikachu clips funny and popular.
The deeper call is that abundant personalization increases the value of common reference points. In Ben’s words, “things that are in common are going to be more and more valuable,” making Disney’s globally recognized characters advantaged even if this particular distribution product fails. Andrew connected the same logic to Warner Bros., saying people are fighting over IP generated decades ago as the company is valued at $80 billion.
Disney’s agreement with OpenAI, paired with a cease-and-desist sent to Google around the same time, also reinforced “Google against the world.” Google owns chips and models and resists deals that do not scale; opposite it sit OpenAI, NVIDIA, and Disney, and Ben said this is where Apple should end up. His hedge remained: “Google might win.”
One reason is Gemini 3 Flash: an “amazing model that’s extremely cheap.” Ben emphasized drastically lower developer API costs and argued that the most revealing benchmark may be the quality of the super-cheap model “that you could actually build a business on,” not only maximum model capability.
5. Netflix is buying against YouTube’s claim on the television
A listener challenged Ben’s attention-market framework, arguing that Netflix also participates in markets for creative talent, subscription video, and network bandwidth, where a Warner Brothers acquisition could reduce competition. The analogy offered was that Amazon’s weakness in streaming would not justify acquiring Walmart.
Ben’s rebuttal began with the shift from scarce physical distribution—film reels, theaters, presses, trucks, and shelf space—to universally accessible internet content. Businesses pay Google and Facebook not to make content available, but to acquire attention; free access consequently expands the competitive field “infinitely.”
On the living-room screen, Netflix therefore competes directly with YouTube, linear TV, Disney+, and even Apple’s aerial screensaver because only one can occupy the display. Ben acknowledged valid monopsony questions and that courts may draw narrower, political market boundaries, but maintained: “YouTube is by far their biggest threat.”
Andrew sharpened the strategic arc: Netflix’s bull case was replacing the cable bundle and becoming all television, while more people now watch YouTube on TV than Netflix on TV. Ben’s formulation: “Netflix was seeking to replace Hollywood for distribution. YouTube is seeking to extinguish it.”
Netflix’s next move is at least 18 podcasts’ video episodes from iHeartMedia and Barstool Sports, publishing exclusively on Netflix from early 2026 and pulling some shows off YouTube. Pardon My Take’s audience was reportedly angry about the change, while Pod Save America’s Tommy Vietor argued it would add reach and actual revenue for growth; the preview ended before Ben delivered his reaction.