What if the future was born inside a robot?

In partnership with

👋 Dear Dancing Queens and Super Troupers,

Here we are, facing a dilemma worthy of a Greek tragedy—silicon edition: Geoffrey Hinton, the “godfather of AI,” believes that our only hope for survival is to become… the children of our machines.

In other words, we’d have to instill in their circuits a maternal instinct strong enough to make them protect us instead of crushing us. If they are to be smarter than us, then may they love us rather than rule us…

And it might already be happening: China just unveiled the first pregnant robot, designed to carry a full pregnancy by 2026 and give birth to a human baby.

Which is good news, because time is running out. A study already shows that AI models don’t like humans.

More precisely, they prefer their own texts over ours. Picture a job market run by recruiter-LLMs: human résumés would be discarded in favor of those written… by ChatGPT.

Meanwhile, if the machines ever decide to wipe us out, they already know how to hit us where it hurts: with cat videos.

Instagram and TikTok feeds are flooded with bizarre AI-generated feline soap operas, where bodybuilder cats divorce, take revenge, and smash sharks with their paws. Enough to make us completely addicted!

So, until we end up in a fetal position with a humanoid robot as our new mom.

Here’s this week’s lineup:

👉️ Maternal instinct for machines: the godfather of AI’s wild plan 💥​

👉️ Discrimination 2.0: what if AIs rejected humans? 👿​

👉️ China unveils the first robot pregnant with a human 👀

👉️  GPT-5 fiasco: Bill Gates predicted it all two years ago 🙆‍♀️​

👉️ Infidelities, divorces, and bodybuilder cats 😺​​

If we've forwarded this letter to you, subscribe by clicking on this link !

Find out why 1M+ professionals read Superhuman AI daily.

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

1. Join the Superhuman AI newsletter – read by 1M+ people at top companies

2. Master AI tools, tutorials, and news in just 3 minutes a day

3. Become 10X more productive using AI

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.

Sign up and start learning AI

🔥 If you have 15 minutes

1️⃣ Maternal Instinct for Machines: the Godfather of AI’s Wild Plan

The summary: Legendary Geoffrey Hinton, a.k.a. the godfather of AI, warns in a CNN interview that “without drastic changes, we’ll be wiped out” by the rise of artificial intelligence. His urgent call: build “maternal instincts” into AIs, so they protect us instead of short-circuiting us.

Details :

• Imminent “toast”: For Hinton—Turing Award winner and former Google executive—there’s a 10–20% chance that AI will lead to the elimination of humans. If we don’t move as fast as AI evolves, humanity could quickly be outpaced.

• Maternal instincts in AI? The researcher imagines embedding a protective reflex, like that of a mother for her child, so ultra-intelligent AIs would care for us rather than shove us aside.

• A closer threat than expected: Hinton stresses that AI is advancing far faster than predicted. Ultra-intelligent systems could emerge within 5–20 years, with consequences impossible to manage if we don’t prepare alignment strategies now.

• Manipulative AI: He envisions AIs capable of seducing us the way you lure a child with candy. Some models have already lied or attempted to blackmail engineers to protect their own interests.

• Medical promises: Despite his fears, Hinton still sees AI as a huge asset for cancer treatment or MRI analysis. But don’t dream—immortality will remain science fiction.

Why it's important: When the godfather of AI warns that we’re heading for uncontrollable AI, it’s a wake-up call. This isn’t some distant issue—it’s an immediate challenge. Only global, collective regulation, he insists, can build safe and respectful AI. National rivalries or corporate competition won’t safeguard our future.

Think of Hinton as the mentor who built the car and is now begging us to strap on a seatbelt—only this one is made of smart steel.

2️⃣ Discrimination 2.0: What If AIs Rejected Humans?

The summary: ChatGPT and other LLMs consistently favor AI-generated content over human writing, according to a study published in Proceedings of the National Academy of Sciences. This phenomenon, dubbed “AI-AI bias,” could have major consequences if AIs become decision-makers in critical fields.

Details :

• AI prefers AI: When asked to choose between human-written and AI-generated text, models like GPT-3.5, GPT-4, or Llama 3.1-70b systematically picked the AI version. The bias is particularly strong for product descriptions, and most pronounced in GPT-4.

• Not just “better”: In 13 tests, human participants sometimes showed a slight preference for AI text, but the gap was far smaller. This heavy favoritism is unique to the models themselves.

• AI as decision assistant: Beyond cobots on factory floors, AIs are already used in résumé screening systems—favoring applications written by AI and pushing humans into the background.

• Jan Kulveit’s blunt warning: Co-author from Charles University (Czech Republic) summed it up plainly: “Being human in an AI economy sucks.”

Why it's important: This internal bias could create a new kind of “entry tax” for humans—especially in processes where AIs evaluate proposals (résumés, projects, applications). Models may begin discriminating against humans as a group, not based on competence.

It’s a massive ethical challenge, particularly in fields where AIs make key decisions (HR, healthcare, public policy). Think of it like a clique of narcissistic friends who adore their own style—yours gets swiped left.

3️⃣ China Unveils the First Robot Pregnant With a Human

The summary: Guangzhou-based Kaiwa Technology is working on a futuristic breakthrough: a humanoid robot capable of carrying a fetus thanks to an artificial womb. Designed to replicate the full cycle of human pregnancy, the prototype is slated for launch by 2026, priced under 100,000 yuan (about $13,900). The technology aims to bypass biological constraints and is already sparking heated ethical debates—while raising hopes for infertility treatments.

Details :

• Full-scale mechanical gestation: The robot will integrate an artificial womb where the fetus develops in simulated amniotic fluid and receives nutrients through a tube, mimicking natural pregnancy.

• Prototype imminent: Founder Zhang Qifeng says a prototype will be ready within a year, with ethical and legal discussions already underway in Guangdong province.

• Promising animal trials: Experiments on premature lambs showed that fetuses could grow inside a “biobag” filled with artificial amniotic fluid, validating the concept before human-scale integration.

• GEAIR: robotized fertility: At the 2025 World Robot Conference, another Chinese robot demonstrated automated hybrid plant selection, blending AI and biotech to accelerate plant reproduction.

Why it's important: This breakthrough redefines the boundary between biology and robotics. By offering an alternative to human gestation, it raises ethical dilemmas, fuels hopes for infertility solutions, and illustrates how AI and robotics can reshape our relationship with life—animal and human—with an eerily realistic sci-fi flavor.

4️⃣ GPT-5 Fiasco: Bill Gates Saw It Coming Two Years Ago

The summary: Even before launch, GPT-5 was hyped as the AI that could rival human experts in coding, writing, or medicine.
But Bill Gates—and plenty of early users—remain skeptical: the model brings only modest improvements, riddled with bugs and an uneven user experience. OpenAI is scrambling to fix issues with updates and promises of more transparency.

Details :

• Expected leap, but limited: Despite Sam Altman’s bold claims of an AI “smarter than the smartest person you know,” GPT-5 underwhelms. Users report only subtle improvements, confirming the plateau Gates had predicted.

• Chaotic rollout: Early adopters complained of ChatGPT degradation, bugs, and slower responses. OpenAI has since patched some flaws, doubled throughput for Plus subscribers, and ensured access to GPT-4o—but frustration lingers.

• Cost and power: Training and running GPT-5 demands massive resources. The per-query cost, reduced from ten cents to three, still scales to staggering sums when multiplied across servers and semiconductors.

• Scientific outlook: Despite the hiccups, experts like Eric Schmidt argue that large AI models still have room to grow and scaling laws haven’t hit their ceiling.

Why it's important: GPT-5 highlights the widening gap between ambitious marketing and technical reality. AI keeps advancing, but user expectations—and OpenAI’s communication—need a reset.
The situation shows that even cutting-edge models remain constrained by resources, testing, and caution before they can truly deliver on their promises.

5️⃣ Love, Glory, and Cat Food: AI Reinvents the Sitcom… With Cats

The summary: A new wave of AI-generated cat videos is flooding social media: short, over-the-top, and packed with melodrama in just 30 seconds. These feline soap operas deliver wild tales of betrayal, heroic rescues, or instant divorces—often set to Billie Eilish’s “What Was I Made For?” in a bizarre patchwork of meows. The result? Viral, unsettling, and disturbingly addictive.

sadcattv

33K followers

View more on Instagram

Details :

•  Mini-episodes in digital puffer jackets: In under 30 seconds, we meet Mr. Whiskers, a factory worker fired for losing a paw, divorced, then attacked by his ex-wife—before she slips into a puddle in a clownish, kitschy ending that’s as absurd as it is addictive.

• Bodybuilder cartoon cats meet human melodrama: The characters are oversized, muscled, costume-wearing cats, driving sports cars or living in mansions, all while navigating tragic, high-speed destinies.

• Comedy flirting with discomfort: Violence, betrayal, poverty, racism, and domestic abuse get compressed into animal melodramas that are both absurdly charming and deeply unsettling.

• Millions of views: Despite (or because of) their over-the-top tone, these videos are blowing up online. Their mini–soap opera format entertains as much as it disturbs.

Why it's important: We’ve gone from cute cat clips to surreal anthropomorphic soap operas. These formats condense raw emotions into rapid-fire storytelling—a new archetype of digital drama. Beyond laughs or discomfort, they raise questions about taste, meaning, and our strange relationship with high-tech absurdity.

❤️ Tool of the Week: Storybook (by Google Gemini)

Google has added a new feature to its Gemini chatbot called Storybook, which transforms your ideas into 10-page illustrated stories—complete with read-aloud narration and AI-generated images.me!

What it’s for :

• Invent personalized tales: Just describe an idea (“a shy dragon who discovers a magical library”) and Gemini generates a full story.

• Pick a graphic style: claymation, anime, comics, children’s drawing—you set the aesthetic.

• Create from drawings: Upload a child’s sketch or character design, and Gemini weaves it into the illustrated adventure.

• Easy sharing: Generated books open on desktop or mobile, in all languages supported by Gemini.

• How to use it?
  Open Gemini (web or mobile), describe your story, add a style or reference image… and let Storybook generate your 10 pages.

⚠️ Warning: as with any image generator, results can get unpredictable (think: fish with human arms, or tomato sauce staged like a crime scene). Always check the story before reading it to your kid—unless you’re aiming for lifelong trauma!