Nvidia’s Microduck Robots Hit 20,000 Preorders

Nvidia’s Microduck Shows How AI Robots Could Reach Consumers

A tiny robot duck just crossed 20,000 preorders after Jensen Huang played with it onstage.

At Nvidia’s GTC Berlin conference, Microduck robots walked, sat, and performed somersaults on voice command, making little electronic noises as the audience watched.

The response was immediate. Preorders passed 20,000 the next day.

What interests me is how little these robots actually need to do to attract buyers. People are already willing to pay for simple, entertaining interactions with physical AI, even before the technology becomes genuinely useful around the house.

There’s another interesting detail: Microduck is open source, so owners can teach it new tricks and share them with others.

That could give the product a longer life than most novelty toys, especially if developers start creating behaviors that ordinary buyers find useful or entertaining.

The bigger question is whether people will keep playing with these robots once the initial excitement wears off.

OpenAI’s AI Agents Find Brain Disease Leads in 32 Minutes

Thousands of AI agents analyzed data from millions of mouse brain cells in just 32 minutes.

OpenAI, PerturbAI, and Scripps Research used GPT-Rosalind agents to examine how gene disruptions affect brain cells, connecting experimental results with existing scientific research.

The agents identified potential links between cholesterol regulation and neurodegenerative diseases, including a new hypothesis involving a gene associated with Parkinson-like symptoms.

One result stands out. Human experts preferred reports from distributed AI agents over single-agent reports 75% of the time.

I think the interesting part is how much existing research scientists may be unable to examine closely because of the sheer volume of data and published papers. These agents can help researchers identify specific questions worth testing while leaving the actual experiments and validation to scientists in the lab.

If the findings hold up, this approach could help research teams investigate more disease mechanisms without spending months reviewing each dataset manually.

But how many of these AI-generated hypotheses will survive real laboratory testing?

Ethan Mollick Thinks We Still Have a Say in AI’s Future

Wharton professor Ethan Mollick has a simple piece of advice for anyone trying to understand AI: start using it.

In his new book, Co-Existence: The Next Phase of AI, released October 20, Mollick argues that AI is becoming harder to avoid and increasingly capable of doing more than conventional software tools.

His earlier book, Co-Intelligence, published in 2024, encouraged people to experiment with AI while carefully checking its work.

That advice still holds.

I think Mollick is right that hands-on experience matters more than following every new AI announcement, because actually sitting at a keyboard and testing these systems reveals both their capabilities and their frustrating mistakes.

There's also something important in his argument that people still have a role in shaping how AI develops and gets used.

As more workplaces adopt AI, people who regularly experiment with it may be better prepared to question how these systems are introduced and where human judgment remains necessary.

But will ordinary users have enough influence over those decisions?

Prohuman team

Free Guides

Explore our free guides and products to get into AI and master it.

All of them are free to access and would stay free for you.

Feeling generous?

You know someone who loves breakthroughs as much as you do.

Share The Prohuman it’s how smart people stay one update ahead.