Update 2/10/25: The pre-sale has closed, but you can sign up to get updates here on the book’s development and get notified when it’s published!

Readers always have a challenge. There are more books than one can ever read in a lifetime. As an author, I've found a similar one: "Why write this book?"

This book is my answer to, "Where should I even start on AI?" Early history, technical underpinnings, business models, geopolitical implications, and things to watch out for in the future are all covered.

The interview series you’ve seen with AI in medicine, AI in electronics design, and very shortly you’ll see in high finance (update: now published!) and more, have just been a few of the people I’ve talked to in researching and writing this book.

Take a look at the very bottom of this post for some snippets to get a flavor of the book.

Get book updates

A “Crowdfunding” Model

Just to be clear: the ship date will be in the fall (August 2025). After purchasing, the books are not coming in a week or so. So why join a pre-sale?

By purchasing at the pre-sale, you will:

Support the book and speed up its publication
Get special acknowledgment in the printed book itself, but also get signed copies (there may be other surprise goodies as well)
Get access to special lists/communities from me, but also the ability to book 1-on-1s for Q&A

My publisher describes their model as a “hybrid” between traditional and self-publishing, where they take some capital risk—by throwing resources at me upfront—but use crowdfunding and book tours to recoup the cost before moving into heavier revision editing, copyediting, layout, audiobook recording… you get the idea.

This model also gets this book out much faster. I got a chance to talk to some literary agents and publishers. For the most part, they all said that it wouldn’t be realistic to expect the lead time would be shorter than 2-3 years in a more traditional model.

What’s the status right now?

The first draft was done in early December. I do have thousands of revisions my editors and I have already identified. Some to tighten up the writing, but also some because events have transpired to generate even better examples to perfectly illustrate the concepts I discuss.

A snippet of the list of revisions — I use Drafts to tag things, and these are just a subset of some of the changes that I haven’t sorted into places in the book yet. The funniest that I *need* to get in there is Hancock, the Stanford misinformation expert who admitted to using ChatGPT (and having hallucinations) in his court statements about misinformation in GPT-4o.

Can I get a sense of what’s in there?

Yes, let’s go on a whirlwind tour of a few book excerpts:

From cultural basis in science function:

“It’s hard to make things inhuman because you’re always making things anthropomorphic, whether you like it or not. A human has no choice when they’re writing,” says Tom Toner during our call. He is an English science fiction writer and author of the Amaranthine Spectrum series, which features an AI as a core character...

To history:

In this case, it’s a story of two boys, both of whom went to the Bronx High School of Science in New York City, just a single grade apart. These two boys would grow up to shape the two great phases of AI before this one. Frank Rosenblatt, who we’ve already met. And Marvin Minsky, described as “the loyal opposition” by Rosenblatt, and who would fight Rosenblatt’s ideas throughout his life despite outliving Rosenblatt by almost five decades—and help bring about a multi-decade winter for the ideas that would eventually become modern neural networks.

To strategic:

Allen also shares on the podcast that the Department of Defense has been “drinking the AI Kool Aid and the autonomous systems Kool Aid really intensely since 2015.” But it was never significant. At least not before the Collaborative Combat Aircraft (CCA) contract—literally, an AI “wingman” for human pilots—was awarded to Anduril and General Atomics in April 2024. Note, not the typical defense contractors like Lockheed and Boeing, who the tech company, Anduril beat.

To technical:

Convolutional neural networks like AlexNet accomplish this by taking many small, overlapping snapshots all over the image (to solve image shifts or mirroring), applying a filter to recognize features (“things”) inside of each tiny snapshot, and then handing the entire thing into the “classic” black-box connected neural network. This is how AI tells the difference between humans, bridges, cars, and cats.

To the future:

We will also need better cooling. We’re already seeing fascinating progress in “sci-fi” technologies like immersion cooling—where entire racks or even data centers are submerged in liquid. Or even take some of that heat and siphon it off into thermal batteries, which can either be turned back into electricity… or simply be used to make steel or other products. After all, 20% of global energy demand is actually just to make heat and we could “recycle” some of it.

Get book updates

Will this be out on Amazon and other bookstores later?

Just to be clear, yes. This isn’t your one-and-only-time to ever buy. This also isn’t the kind of crowdfunding promising that if you jump in now, you’ll get X% off! (Those often don’t go particularly well anyway…)

The explicit reason to buy now is to both support getting the book out, get some behind-the-scenes as it comes together (and even give feedback), and—I know this has been important for some folks—to get more interaction and thoughts from me about things within your specific industry.

You’ll see that some of those packages support that (and a few are already taken).

You’ll also get some cool acknowledgments/credits printed too, and one (or more) signed copies!

Please share!

Even if you want to wait, I’d love for you to share and get the word about the book out—whether or not those friends want to pick it up in stores, or if they want a fancy signed early copy!

Get book updates

My Book's Pre-Sale is Live!