Tony Fadell, speaking at the inaugural MIT Future Fest, opened with a slide showing three discontinued AI gadgets: the Rabbit R1, the Humane AI pin, and the Limitless pendant. He called them “Gen 1” AI products and revealed that the companies behind them had reached out for his help, but he declined. Fadell, known as the “father of the iPod,” a co-creator of the iPhone, and founder of Nest, argued that these devices failed because they did not solve a real need. “In each of these cases, each one of them didn’t meet any kind of need — it was just interesting technology for geeks,” he said.

Fadell pointed to a deeper problem: most people have never had a personal assistant and therefore do not know how to use or trust one. He noted that less than 0.01% of the world population has ever had a human assistant do anything for them. When companies say they want an assistant, they are speaking from a narrow perspective. For the other 99.99%, trusting an AI with sensitive data is not intuitive. Fadell shared his own experience of hiring an assistant, explaining that it took years to understand how to best utilize one and to trust the person with sensitive data, bank accounts, and meeting scheduling. He acknowledged that there is a long path before general consumers and businesses are ready for AI agents.

Trust and security are central to the future of AI assistants. Fadell said, “Trust and safety are going to be paramount with anything that we turn over to some kind of intelligence.” He identified Apple as one of the few companies capable of doing this well, citing its hardware, chips, and ecosystem. However, he noted Apple does not have all the AI components, pointing out that the new Siri AI runs on custom-built versions of Google’s Gemini. Even with that gap, Fadell believes Apple’s privacy reputation gives it an advantage with consumers.

Fadell also predicted that a successful AI agent will need to operate on-device only, both for privacy and for lightweight performance. He dismissed the idea that data centers will conquer the world, saying he has seen similar claims during the internet days. Devices now have enough power and compute, and they remain battery operated. Keeping sensitive data on-device, rather than transmitting it through the cloud, is how Apple has built trust for features like Face ID.

Fadell theorized that companies like Meta and OpenAI are building dedicated gadgets because they cannot access the full set of sensors on users’ phones. When an app asks for video, audio, GPS, and other permissions, the list grows quickly. Instead of dealing with those permission prompts, these companies build standalone devices with all the sensors built in, then connect to a phone via Bluetooth or Wi-Fi and back up to the web through a 5G network, giving them direct access to sensor data.

Despite his Apple history, Fadell is not uncritical of his former employer. When an audience member asked about luck in product-market fit, he noted that startups get one shot, unlike Apple, which can absorb failures like the Vision Pro. For startups, a failed product can end the company, making the manufacturing of luck far more difficult.