I Am Not a Robot: My Year Using AI to Do (Almost) Everything

I found this book from an episode of the Decoder Podcast. The premise is that Joanna Stern used AI, LLMs, and robots for many of her tasks and jobs for 2025. The book goes through her experiences, what the tools did well, what they struggled with, and what they simply failed miserably at.

I appreciated her humor in her writing and talking about some of her experiences, both good and bad. As you read the book, she does a good job of separating the different chapters or sections by the type of tool or AI assistant she's trying out. She's honest about what they can and cannot do, as well as what the tools are. She's also not trying to sugarcoat things either as evidenced by her mentioning of the power requirements of the data centers powering all of this.

In northern Virgian, Dominion Energy, one of the main power providers has said that contracts for data cetner capacity jumped from 21 gigawatts in mid-2024 to nearly 40 gigawatts by the end of the year. For reference, 1 gigawatt can power about 750,000 homes for a year. And Sam Altman is out there saying OpenAI might need 10 gigawatts just to train AI systems that can tutor every kid or cure cancer. That's one company asking for the power of seven and a half million homes. Take a minute to think about that, maybe while sitting with the lights off.

Towards the end of the book, there was this line, that still sticks with me. It's one of the reasons why AI content just rubs me the wrong way. The human aspect is why I read books and good journalism.

Sending AI Joanna to internal meetings to ask questions about font sizes and data metrics, sure. But the core of this work, talking to people, seeing things through my own eyes, asking weird unscripted questions, is where ideas are born. The hundreds of real conversations I had for this book are what gave it shape and soul.

In the end, she's probably more favorable towards what AI can do for us than I am, but with a much more realistic bend than most.