NEWS YOU CAN’T MISS

China's Secret to Winning the Robot Race

Before robots can help out in homes or hospitals, they need to learn how the world works—and one way to do that is by watching us.

At a new training center in Shanghai, over 100 humanoid robots are doing just that. They’re being taught to fold clothes, pick up objects, and clean machines by observing people repeat those tasks over and over again. It’s backed by the Chinese government, and is part of the country’s broader push to lead in humanoid robotics—a field it sees as strategically vital for the future of manufacturing and healthcare.

It might sound simple, but there’s a much bigger goal: building robots that can learn anything, not just one pre-programmed action. The center is trying to solve a few key problems:

• Different robots speak different “data languages”, which makes it hard to share what one has learned with another. This project is creating a shared dataset they can all learn from.

• The focus is on basics—grasping, placing, sorting—because those are the building blocks for more complex tasks.

• The long-term goal is a shared “robot brain”, so machines from different companies can learn, adapt, and even work together.

With up to 50,000 motion data entries collected each day, it’s a big step toward smarter, more useful robots in homes, hospitals, and factories.

This is a big deal. Most AI progress so far has come from massive datasets of text, images, or code. But humanoid robots need something different: millions of real-world movements to learn how to function in physical space.

This center is one of the first serious attempts to create that kind of shared motion dataset at scale. If it works, it won’t just improve an industry or two—it could lay the foundation for embodied AI the same way large language models did for chatbots. That’s a major shift in how we teach machines to think and move. 😎

Outsourcing Love: When a Bot Replaces the Family Call

For busy families, keeping in touch with aging parents can be tough. A new service called inTouch offers a solution: AI-powered phone calls designed to provide companionship and mental stimulation for older adults.

How it works: you sign up for a subscription, and your parent receives regular calls on their landline or mobile. The AI system holds conversations on a range of topics, from family memories to hobbies and brain teasers. After each call, you get a brief summary, including how your parent responded and what they discussed.

The idea is to reduce loneliness and support cognitive health through daily social interaction—especially for those who live alone or experience memory loss. inTouch also includes mood tracking and a mobile app, giving family members insight into how their loved ones are doing.

Honestly, I heard the demo on the website and I felt really sad when the AI asked the Dad if he had message for his daughter, and he said, “Yes, please tell Sarah that I love her.” But is this better than never calling your parents because you are running between work and kids and deadlines?

I don’t know… it just made me sad. What do you think?

Faking AI Automation? That’s Fraud, Says the Justice Department

US-based shopping app Nate promised users a seamless, AI-powered checkout. Just tap “buy,” and its proprietary tech would handle the rest. The pitch helped raise over $50 million from top VCs like Coatue and Forerunner Ventures.

But the automation was a façade. According to a federal indictment, Nate relied almost entirely on human workers in the Philippines and Romania to process orders. The actual AI usage? “Effectively 0%.”

The truth began to unravel in 2022, when The Information revealed that Nate’s so-called AI was, in reality, hundreds of human workers in the Philippines manually placing orders.

Whistleblowers and former employees said most of the company had no idea how little automation existed, as access to performance dashboards was tightly controlled. When some questioned the lack of transparency, CEO Albert Saniger reportedly dismissed their concerns, calling the data a “trade secret.”

Following the exposé, Nate struggled to raise additional funding and quietly shut down in 2023. A subsequent FBI and DOJ investigation confirmed that the app was “a scheme filled with smoke and mirrors.”

Founder Albert Saniger now faces wire and securities fraud charges, each carrying a maximum 20-year sentence.

Nate’s story is a cautionary tale about AI-washing—using the hype around artificial intelligence to raise money without real tech behind it. It erodes trust, misleads investors, and makes it harder for truly innovative startups to be taken seriously.

“Big Trouble” for Consultants Who Don’t Adapt

Uber cofounder Travis Kalanick thinks AI is coming for consultants—and not all of them will make the cut.

Speaking on a recent podcast with Peter H. Diamandis (co-founder of Singularity University), Kalanick said consultants who mostly execute instructions or handle repetitive tasks are in “big trouble.” In a future driven by AI agents, he joked, you might just “push a button, get a consultant.”

But it’s not all bad news. The consultants who thrive, he said, will be the ones building tools—not just using them. In his words: if you’re the person “that puts the things together that replaces the consultant, maybe you got some stuff.”

It’s happening already; firms like Deloitte and EY are already deploying AI agents to handle client work, from data analysis to tax compliance. Deloitte’s new Zora platform offers a roster of “intelligent digital workers,” and EY is rolling out 150 AI tax agents to support 80,000 employees.

Watch the episode here - Kalanick shares these insights around minute 48 to 50.