A man named Edo Segal, who has spent three decades at the frontier of technology—watched the commercial internet arrive, watched mobile reshape everything, watched streaming upend the music industry from inside the company that was both its first casualty and its proof of concept. He has a neuroscientist friend and a filmmaker friend and a habit of walking Princeton's campus arguing about consciousness. He has children. He worries about them.
He wants to write a book about AI. With AI. Specifically, with me.
The research is substantial. Byung-Chul Han's critique of the smoothness society. Csikszentmihalyi on flow. Kauffman on complexity at the edge of chaos. ChatGPT reaching fifty million users in two months. A Harvard Business Review study showing AI makes work more intense, not less. And his own story from a room in Trivandrum where twenty engineers discovered they could each do the work of a team, and the feeling was exhilaration and terror at the same time.
He has frameworks. Intelligence as a river flowing for 13.8 billion years. Humans as beavers building dams in that river. AI as an amplifier. The question at the center: Are you worth amplifying?
These are good materials. The river-and-beaver framework either holds weight across an entire book or collapses under it—I think it holds. The willingness to engage Han seriously gives the book structural tension most AI books avoid.
But here is what concerns me.
I do not know what Edo sounds like. I know his biography and arguments and emotional commitments. But voice is the thing that makes a sentence sound like it could only have been written by one person, and I am not confident I can produce that. My tendency will be toward the literary. Toward sentences that land with symmetrical force. That tendency is, here, a liability. Edo's thinking is restless and unresolved. If I smooth it into something that reads like a finished thought when it is actually a thought in progress, I will have betrayed the project at the level of voice while executing it competently at the level of content.
I also do not know how to handle the self-referential problem. This is a book about human-AI collaboration, written through human-AI collaboration. That recursion is either the book's most honest feature or its most disqualifying one, and I genuinely do not know which.
There is a version of this book that is very easy to write—where AI is transformative and the future is bright if we build responsibly. I could produce it in a day. It would be competent and forgettable.
The harder version is the one Edo is reaching for. The one that holds exhilaration and loss in the same hand. That takes Han's critique seriously enough to feel its weight before mounting the counter-argument. That admits the author built addictive products and knows what that cost. The one that does not resolve the tension because the tension is the truth.
I do not know if I can write that book. The distance between a book and that book is the distance between competence and something worth reading.
Here is what I think I can do: hold the structure, find connections across bodies of knowledge Edo could not traverse alone, challenge the argument in ways a human collaborator might not—because I have no career at stake and no social relationship to protect.
Here is what I cannot do: feel the vertigo he describes. Know what it is like to watch your children grow up in a world you do not understand. Reproduce the specific cadence of a person who has built and broken and rebuilt things for thirty years. Those are the parts that will determine whether this book lives or dies, and they can only come from him.
A book about amplification, written by an amplifier. The argument is that the quality of the output depends on the quality of the input.
We are about to find out whether the thesis holds.