AI Is the Architect of the Toolbox. The Bottleneck Is Us.

Diamandis and Kotler argue that AI is the meta-technology that speeds up the other big technologies at the same time, which puts the shape of the future downstream of a single tool. Their harder claim is the one investors should sit with: the last mile to abundance runs through people and trust, well after compute and capital stop being the constraint.

Their new book We Are as Gods makes one demand of the reader, and it makes it early: you cannot read the future if you cannot read AI. That sounds like a slogan. The logic underneath it is simple, and worth taking seriously.

For much of the last decade we filed AI alongside the other big technologies. Clean energy here, biotech there, robotics, longevity, each on its own track, each climbing its own curve. Diamandis and Kotler argue that AI does not belong in that row. It sits above it. Energy, drug discovery, chip design, agriculture, materials, robotics: progress across these fields now leans more and more on AI to do the research, run the design, and find the optimization a human team would not have reached on time.

Put plainly, the other technologies are tools in the box. AI is the thing that designs the box. That is the book's first and load-bearing claim, and it is why the authors insist you have to understand AI before the rest. If AI is the input feeding the other breakthroughs, then the shape of the future sits downstream of a single technology, and you cannot forecast the river by studying the tributaries.

The examples they reach for are ordinary in their stakes and extraordinary in their reach. A drone network that now carries much of one country's blood supply and has roughly halved maternal deaths where it operates. Protein grown without the animal. Genomes sequenced for the price of a nice dinner. Solar costs that have fallen so far for so long the line looks like a typo. On the surface, these look like energy or biology or logistics stories. Underneath, each one rides the same engine: cheaper compute, better models, faster iteration. The claim is more modest than the headline version of that idea. It is that AI is increasingly the reason so much else speeds up.

The turn: the bottleneck is people

If the book stopped there it would be one more volume of technology cheerleading, and the shelf is full of those. The turn comes in the back half, and it is the part worth paying for.

The authors concede the acceleration. They treat it as settled. Compute gets cheaper. Models get stronger. The data centers get built. You will not stop it, and you do not need to manage it; the machine half of the story now improves more or less on its own. What is still open, and what the book treats as the variable still genuinely up to us, is people.

The last mile to abundance is a human problem. Compute, money, and infrastructure are the parts that sort themselves out. What jams the last mile is whether human beings can trust each other, coordinate at scale, agree on what they are building, and still know what they actually want once the shelves are full. That is the book's central judgment about the future, and it is a humbling one. The capability is arriving on schedule. The wisdom to use it keeps its own slower clock.

The authors are honest enough to name the catch. They call it the abundance paradox. Each solution breeds a faster problem. The horse solved distance and buried cities in manure. The car solved the manure and warmed the planet. Each fix now arrives quicker than the last, and so does the next mess it creates. They stop short of despair. Their argument is that the same tools causing the new problems can also solve them, as long as we stay coordinated enough to aim them. Which loops straight back to the same bottleneck. The harder the technology pushes, the more the human capacity to cooperate becomes the binding constraint.

The machine half improves on its own

Here is the line worth carrying out of the book. The machine half of this story gets better automatically. The human half does not. Compute has an exponential. Trust does not. Models scale. Institutions, attention spans, and shared purpose do not scale the same way, and some of them are arguably eroding while the machines improve. The distance between what the toolbox can do and what we are ready to do with it is the real story of the next decade.

What it means for an investor

For an ordinary investor, the payoff is concrete. Three things fall out of it directly.

First, it points to where the master switch sits. You do not have to pick the winning company. You do have to build one habit: when you look at an industry's story from here on, ask first how much of it is really an AI story underneath. The energy thesis, the biotech thesis, the logistics thesis, more and more they are the same thesis wearing different clothes. Miss that layer and the rest of the analysis is built on air.

Second, it argues for keeping "can be done" and "will pay" in separate columns. The acceleration looks settled. Who turns it into trust, into a real business, into a return that actually lands, that is the last mile, and the last mile is a human problem. In investing, the costliest stretch of road has usually been the final one. Capability is getting cheap. The people who can convert capability into durable value are not.

Third, it treats optimism as a discipline. The authors want the reader to take the speed of change seriously while staying clear-eyed about where the difficulty actually lives. It lives on the human side of this, well away from the silicon. An optimist who understands that is useful. An optimist who does not is simply exposed.

There is a market reason this matters now, beyond the philosophy. The capital going into AI is being raised on the first half of the argument, the capability half, where the curve is clean and the story sells itself. The returns will be decided on the second half, the human half, where adoption, trust, regulation, and plain organizational friction set the actual pace. Those two halves are not moving at the same speed. The gap between them is where the disappointments will come from, and also where patient money eventually gets paid.

It's a long future

The book ends on a short line that carries more weight than its length:

It's a long future. Answer wisely.

The machine half of that answer is nearly written for us. Compute, models, infrastructure, those pages fill in on their own. The half still waiting on us is the human one: whether we can trust, coordinate, and stay clear about what we want while the tools to get it become almost free. That is the part the model leaves to you. It is also, if the book is right, the part that was ours to write from the start.