👋 Dear Dancing Queens and Super Troupers,
This week, Disney pulled out the checkbook.
Not a small, discreet scribble in the margins — no. One billion dollars placed squarely on the table for OpenAI.
A clean, massive, almost theatrical move. And suddenly, generative AI is no longer just the obsession of geeks, hungry startups, or NDA-bound research labs. It officially becomes a studio business — a matter of licenses, iconic characters, and collective imagination.
Because when Disney takes a seat at OpenAI’s table, it’s not to “test a POC.”
It’s to redefine who will create stories tomorrow, how they will circulate, and — most importantly — who gets to hold the controller.
Mickey, Iron Man, Buzz Lightyear… tomorrow, generated, remixed, animated by AI models accessible to the public. Hollywood is no longer asking whether AI will transform creativity. It’s negotiating its place inside it.
And it’s no coincidence that this announcement lands in the middle of such a week — one packed with contradictory signals.
On one side, OpenAI signs with Disney and reinforces its industrial aura.
On the other, a researcher slams the door, accusing the company of smoothing over the truth, while users grumble about a ChatGPT 5.2 perceived as colder, more rigid, more “corporate.”
As if, the more power and prestigious partners AI gains, the more it loses a bit of its original spontaneity.
Add to that the announcement of an upcoming “adult mode” planned for 2026 — a reminder that we are now talking about a tool used by hundreds of millions of people, with very different expectations… and not always compatible with a one-size-fits-all sandbox morality.
And let’s not forget the head of Hinge stepping down to launch an AI-powered dating app, proof that even our love stories are now considered optimizable by machines.
Here’s this week’s lineup :
👉 When Disney embraces OpenAI, Hollywood rewrites the script 🎬
👉 We’re hiding the truth”: a researcher quits OpenAI and speaks out 🕵️♂️
👉 Fewer filters, more freedom: ChatGPT comes of age 🚀
👉 ChatGPT 5.2 feels colder: did OpenAI miss the tone? ❄️
👉 Overtone: Hinge’s former CEO goes all-in on AI 🎯
If we've forwarded this letter to you, subscribe by clicking on this link !
⚡ If you have 1 minute
Disney puts a billion dollars on the table and partners with OpenAI to leverage its licenses (Disney, Marvel, Star Wars…) in video generation via Sora. This is not a test — it’s a massive strategic bet on AI as the future engine of cultural creation. Hollywood is no longer watching AI from a distance; it’s signing deals.
A former researcher claims OpenAI is slowing down the publication of research highlighting the potential negative effects of AI, particularly economic ones. The departure reignites the debate around transparency and the company’s commercial pivot. Behind the polished product façade, the “open” culture appears to be under serious strain…
OpenAI is preparing a less restrictive version of ChatGPT, reserved for adults, with freer, less filtered conversations. The launch has been pushed back to 2026 to secure reliable age verification. This isn’t just about NSFW content — it’s an acknowledgment that mass-market AI can’t speak to everyone like they’re in middle school.
Early users describe ChatGPT 5.2 as colder, more rigid, and less pleasant than previous versions. Technically more stable, but less emotionally engaging. A reminder that in consumer AI, perceived experience sometimes matters just as much as actual performance.
Justin McLeod leaves Hinge to found Overtone, a dating app built around AI and voice, with backing from Match Group. The goal: move beyond swiping and create deeper connections. AI continues to invade our lives… all the way into our romantic relationships.
🔥 If you have 15 minutes
1️⃣ When Disney Embraces OpenAI, Hollywood Rewrites the Script
The summary : Disney is clearly accelerating its technological shift. The entertainment giant is investing $1 billion in OpenAI. The agreement allows the video tool Sora to leverage a carefully controlled portion of Disney’s catalog — more than 200 iconic characters and universes.
This alliance outlines a new framework between cinema and automated video generation, while protecting actors and intellectual property. In the background, Disney is simultaneously taking a harder stance against Google over the use of copyrighted content.
Details :
A major strategic investment: Disney injects $1 billion into OpenAI. This project confirms its ambition to play a key role in generative creation and to integrate these technologies into its digital services.
Sora enriched with Disney’s catalog: More than 200 iconic characters and settings join Sora, including Mickey Mouse, Cinderella, Black Panther, and Luke Skywalker — without using actors’ faces or voices.
Videos designed for Disney+: Some Sora-generated creations are expected to be distributed on Disney+, a strong signal of the importance assigned to the project.
A reassuring approach for studios: By excluding actors’ likenesses and voices, Disney limits concerns related to rights and licensing.
Google under legal pressure: At the same time, Disney’s lawyers have reportedly sent a letter to Google accusing it of allegedly using protected content to train AI models, according to Variety.
Bob Iger as a discreet conductor: Disney CEO Bob Iger maintains close ties with Thrive Capital, a key OpenAI investor, and has held a stake in its management company since 2023.
Why it's important : This deal charts a more measured path between Hollywood and AI. Disney shows that tightly framed collaboration can fuel innovation without sacrificing original works, while redefining the rules of global digital creation.
2️⃣ “We’re Hiding the Truth”: A Researcher Quits OpenAI and Speaks Out
The summary : A quiet unease is spreading inside OpenAI. According to Wired, at least two economics researchers have left the company. They denounce a growing reluctance to publish studies suggesting that AI could weaken the economy. Among them is Tom Cunningham, who accuses the research team of shifting from an analytical role toward message-driven communication.
In the background, OpenAI continues its transformation. Founded in 2016 as a nonprofit organization, the company has become a public-benefit, for-profit player engaged in massive financial bets.
Details :
A resignation that makes noise: Tom Cunningham, an economics researcher, leaves OpenAI and writes internally that economic research now resembles a promotional tool rather than independent exploration.
A hardened official line: Jason Kwon, Chief Strategy Officer, explains that the company must prioritize “solutions” rather than publishing research on complex or uncomfortable topics.
A wave of departures: According to Wired, at least two members of the economic research team have left, frustrated by a climate perceived as hostile to analyses critical of AI.
An openly embraced transformation: OpenAI, founded in 2016 as an open-source lab, now relies on proprietary models. The firm is targeting an IPO valued at $1 trillion.
Staggering financial stakes: OpenAI has raised billions, secured up to $100 billion from an AI chip manufacturer, and committed as much as $250 billion to Microsoft for Azure infrastructure.
Research under scrutiny: A report by Aaron Chatterji highlights ChatGPT’s global use for productivity, while studies praising AI’s benefits continue to multiply.
A broader unease: Before Cunningham, several researchers had already left OpenAI over ethical, safety, or scientific disagreements.
Why it's important : This case highlights the central tension of modern AI: can a technology still be openly criticized when it is worth hundreds of billions of dollars? At OpenAI, research now appears to be walking a tightrope, suspended between scientific truth and economic imperatives.
3️⃣ Fewer Filters, More Freedom: ChatGPT Comes of Age
The summary : OpenAI is finally stepping out of ambiguity. The company announces it is targeting Q1 2026 for the launch of ChatGPT’s adult mode — a timeline dependent on one key factor: the reliability of its age-prediction systems.
Long mentioned by Sam Altman without a clear schedule, this evolution is now official, deliberate, and structured. The goal is to enable adult-oriented use cases without ever exposing minors. All of this unfolds within a regulatory environment that is tightening worldwide.
Details :
A date finally set: OpenAI confirms, via Fidji Simo, a launch planned for the first quarter of 2026. This is the first publicly announced time window for ChatGPT’s adult mode.
An openly acknowledged strategic shift: Previously cautious, the company now recognizes growing demand for adult-only content. The announcement repositions ChatGPT against its direct competitors
Competition in sight: Chatbots like Grok, developed by xAI, already offer adult features. OpenAI is observing and analyzing — but refuses to rush.
Age prediction as the cornerstone: Before activation, OpenAI wants to secure a model capable of reliably identifying users under 18. Tests are underway in several countries to assess its accuracy.
Zero tolerance for error: A single misclassification would pose a major risk. OpenAI prefers to delay the launch rather than undermine public trust.
Rising regulatory pressure: Europe and the United States are strengthening age-verification requirements. Adult mode will need to comply from day one.
Strictly framed use cases: Sam Altman mentions sexual conversations between verified adults, restricted to clearly identified profiles.
Why it's important : Because ChatGPT is no longer just experimenting. By setting a 2026 target, OpenAI turns a sensitive topic into a structured project. A cautious but decisive step for a tool that has gone mainstream — and is now, quite literally, growing up.
4️⃣ ChatGPT 5.2 Feels Colder: Did OpenAI Miss the Tone?
The summary : Barely released, ChatGPT 5.2 has already triggered a wave of unexpected criticism. OpenAI presents it as “the most intelligent consumer model in the world,” according to Sam Altman.
Yet on Reddit, several long-time users describe an experience they find bland, overly formal, or even less inspiring than ChatGPT 5.1. The release comes at a tense moment, as Google impresses with Gemini 3. In less than 24 hours, the debate intensifies: genuine progress or a temporary misstep?
Details :
A very ambitious promise: OpenAI claims that ChatGPT 5.2 outperforms all models available to the general public — a bold statement at the very moment Gemini 3 is gaining momentum.
A frosty reception on Reddit: On the ChatGPT subreddit, a thread titled “So, what do you think of version 5.2?” quickly attracted numerous negative reactions on day one.
Blunt criticism: User AsturiusMatamoros sums up the unease by describing the model as “too institutional” and “a step back compared to version 5.1.”
A tone seen as too robotic: Several posts denounce a chatbot perceived as rigid and unengaging, compared to “a corporate robot.”
A loud minority: TechRadar nonetheless reminds readers that ChatGPT has roughly 800 million weekly users, and that dissatisfied users often speak louder than satisfied ones.
A verdict still premature: Less than 24 hours after launch, even TechRadar admits it is still too early to objectively assess the real progress of ChatGPT 5.2.
Why it's important : This immediate backlash highlights the extreme pressure weighing on every new AI release. At a time when Gemini 3 is redefining expectations, ChatGPT 5.2 is playing for high stakes: prove itself over time — or adjust quickly to prevent disappointment from taking root.
5️⃣ Overtone: When Hinge’s Former CEO Goes All In on AI
The summary : Justin McLeod, founder and CEO of Hinge, is stepping away from the helm to launch Overtone, a new dating app powered by artificial intelligence. Financially backed by Match Group, the project aims to reinvent romantic connections.
By relying on voice and AI-driven tools, the platform seeks to foster deeper, more authentic exchanges. Meanwhile, Hinge changes leadership but stays the course on AI, with already measurable results.
Details :
Justin McLeod switches arenas: Founder of Hinge in 2011, Justin McLeod officially leaves his position to create Overtone. He will nonetheless remain a consultant at Hinge until March, ensuring a smooth transition.
Overtone, a project incubated at Hinge: Developed over one year by a dedicated team, Overtone focuses on AI and voice-based systems. These tools promote more thoughtful and personal connections, far removed from endless profile swiping.
Match Group bets big: Owner of Hinge, Tinder, and OkCupid, Match Group is financing Overtone at the pre-seed stage and plans to take a “substantial stake,” according to its official statement.
A dating app market under pressure: Tinder has recorded nine consecutive quarters of declining paid subscribers. To reignite growth, AI is becoming the primary weapon against user fatigue, particularly among Generation Z.
Hinge continues its transformation: Jackie Jantos, now CEO, notes that the AI-powered recommendation feature launched in March increased matches and contact exchanges by 15% in the first quarter. Hinge is still targeting $1 billion in revenue by 2027.
Why it's important : AI is no longer just optimizing matches — it is reshaping the entire dating experience. Between promises of more genuine connections and concerns over personal data, online dating is entering a new era, more technological than ever.
❤️ Tool of the Week : Google Disco Turns Your Tabs into Apps
With Disco, Google is testing a simple yet radical idea: stop treating the web as a series of passive tabs. Powered by Gemini 3, GenTabs turns your browsing session into temporary applications generated on the fly, directly from what you’re already viewing.
What is it for?
Turning a research session into an interactive tool (visualization, summarization, planning)
Creating custom mini-apps without coding, simply by describing what you want
Leveraging multiple tabs at once to structure a complex task
Refining and evolving the generated app using natural language
Maintaining a clear link to the original source material used
How to use it?
You browse as usual. Disco analyzes your open tabs. GenTabs suggests a relevant app — or you explicitly request one via prompt. The app appears, adapts to your instructions, then disappears once the task is complete.
💙 Video of the Week : Pixar, AI, and a Taste for Chaos
On Sora, OpenAI’s video tool, users are already hijacking Disney and Pixar aesthetics to generate AI-made trailers dripping with absolute cynicism: dark humor, shattered taboos, racist or sordid provocations. A disturbing plunge into what pop culture becomes when AI removes the brakes… and exposes our collective impulses.
OpenAI triggers Code Red: normal reaction or full-on panic?
