And Then AI Showed Up
This is Field Note 001, Part 2. Part 1 focused on a letter one of my professors left in my mailbox after the first semester of my PhD program, the pattern I started noticing everywhere afterward, and the puzzle I carried into my dissertation research, long before AI entered the conversation. MISS PART 1? READ IT HERE
In late 2019, I began pulling together an event study on AI acquisitions. I had no idea at the time that I would reorient my life around what I found. Real quick, an event study asks what investors think an announcement is worth. You estimate what a firm’s stock should have returned given how the market moved, compare that to what it actually returned in the days around the announcement, and then average that difference across every firm in the sample. If the average is reliably different from zero, you have evidence about the category of announcement rather than a story about one company.
AI was not my area. The marketing-finance interface was, and the event study was the tool I used. I came up through a finance degree and then a marketing PhD, which is how I ended up spending my time watching what investors made of what companies did. I had a tendency as a scholar to investigate areas that had real world momentum, because they were highly citable and genuinely exciting to me. But the thing that stopped me in my tracks after the first round of data collection is something everybody can see plainly today. A small number of names kept showing up as the repeat buyers, Facebook, Amazon, Microsoft, Google and Apple, and for many of the deals the price was sealed. I remember sitting there thinking, what the hell is going on? I was too ignorant of the space to understand what I was actually looking at.
The event study ended up with 204 acquisition announcements made by 48 publicly traded, U.S. firms, with those Big 5 doing a disproportionate share of the buying. The market liked them. Abnormal returns were positive and significant around the announcement. Whatever was being consolidated, and whatever it cost, investors approved.
I didn’t know it then, but that event study was the beginning of me waking the hell up. It made me curious, inspired, and frankly a little nervous the more I learned about the relationship between AI and power. Nobody, in my circle at least, saw it the same way. But I was obsessed. It was all I could talk about. I read everything I could get my hands on, and the whole time it felt like something enormous was happening that hardly anyone around me seemed to notice. I started to wonder if I was the one losing it. In fact, when I moved to Alabama, my department chair perpetuated that insecurity, comparing the excitement around AI to the hype that had been around blockchain a few years before. It's hard to hold on to your confidence when the person who runs your department tells you the thing you can't stop thinking about isn't a big deal.
But I kept going anyway. I kept pushing because of what happened when I stopped asking these systems for answers and started building the conditions a good answer requires. I would ask what fields of expertise a question like mine even required, and then work through them one at a time until I understood enough to use them. I would show the system what elite looked like at the highest level I could imagine, feed it examples of where we were trying to go, and flood it with context from inside the situation and from outside it, until it could see what I was looking at. Then it would mirror the thing back to me, and I would argue with it, and we would go again, and again, until what came back was something I could examine, hand to another person, and trust.
What used to take a team and a long stretch of time was happening for me in an afternoon. The distance between imagining something and planning it, prototyping it, testing it, and explaining it to someone else had started to shrink. I want to be careful here, because none of the hard parts went anywhere. Material constraints, institutional politics, money, law, infrastructure, and the need for human judgment were all still sitting exactly where they were. What AI did change was the economics of turning a half-formed idea into an intention, and then into a first draft of a new reality, and the version I kept reaching for was the most ambitious and meaningful one I could think of because why wouldn't you if you could?
Which is when the big question arrived, though I had been staring at it in that acquisition data years before any of this had a consumer interface. If a first draft of reality is suddenly cheap to make, who gets to write it? Who gets to describe futures others have to live inside?
That should have felt like hope, and for me it was. But hope was not what I was hearing most often. What I was hearing was efficiency. Save time. Produce more. Do the same thing faster. And I found that genuinely depressing. We had been handed a tool that could help us question, imagine, prototype, and build — and the first instinct in so many places was to point it at an objective nobody had paused to examine. Be more efficient at extraction. Get more precise at targeting. Maximize the thing we were already maximizing, only faster.
If you take a system organized around an old objective and make it dramatically more efficient and precise, you don't automatically get a better system. What you actually get is the current system, sharpened. The question that stopped me cold was simple: what happens to actual human beings when we become more and more precise at objectives that were never designed around their flourishing?
What happens to actual human beings when we become more and more precise at objectives that were never designed around their flourishing?
The danger I could see most clearly was intelligence aimed at an objective no one had paused to question.
Raising the Floor
The same realization contained another possibility, and it is the reason I began building what I am building. If AI could shorten parts of the distance between intention and reality, then it could help us pursue something better than a more efficient version of the old thing. It could help us describe a company where people come out more whole and more capable, where customers have reason to trust it, and where the system gets better over the long run, and then help us take the first real steps toward building it.
So the floor should come up. That is my hope in one image. If the efficiency gains are real — and they are — they should not merely bloat the old objective. They should raise the floor for the people inside these systems, for the people these systems decide things about, and for all of us who have to live inside whatever gets built.
And that's where one of my biggest epiphanies to date came from, the one I haven't been able to put down since.
The AI future has been branded. The human dream has not.
We have confident, well-funded, relentlessly repeated language for what intelligence can do — fast, disruptive, productive, efficient, inevitable. That language travels because people with money and platforms make sure it travels into keynotes, board decks, university strategies, government rooms, and the low-grade anxiety many people carry now: wait, am I already behind something I don't even understand?
I mean, what's all this intelligence for anyway? What kind of life is it meant to serve? What kind of people is it helping us become?
The actual human aspiration underneath all of this hasn't received the same imagination, creativity, force, or seriousness. I mean, what's all this intelligence for anyway? What kind of life is it meant to serve? What kind of people is it helping us become? Narrative isn't the only force that determines the future; capital, law, infrastructure, capability, and geopolitics matter enormously. But the futures we make culturally compelling shape what people fund, tolerate, demand, and even attempt. And every day the human purpose goes unspoken, acceleration without purpose takes more ground by default. I can't accept that. Mostly because I'm a father. So the human dream, the way I had begun saying it, was almost embarrassingly simple. Pain at scale should go down. Possibility at scale should go up.
Pain at scale should go down. Possibility at scale should go up.
That's it. It sounds too simple to carry this much, but I don't think it's a small thing. I think it might be the whole thing, and it became the design brief for everything I have built since. The messier realization underneath it was that a new tool doesn't change the objective it is pointed at, so someone still has to ask what the power is for. And the more I looked, the fewer people I could find who were in any position to answer that.
Finding the Language
I didn't yet have all the language for Human Readiness. I was beginning to see a gap between the power becoming available and the human capacity around it, which I have come to think of as three things. Whether an institution can see itself honestly, including what its own people already know. Whether it can imagine a future worth building. And whether it can move with intention instead of by default. I started calling the distance between that power and those capacities the Human Readiness Gap.
I was also beginning to think of AI as a kind of mirror before an amplifier — a way of making patterns, incentives, exclusions, and assumptions easier to see. I would later have to complicate that claim, because a mirror can distort, and the first part of the founding note says how. But even this early, I could see that the first question about AI was never the platform. It was what kind of institution the platform was entering, what that institution was organized around, and whose reality could reach the people making the decisions.
And that's where I need to stop, because the next thing I learned was harder. Seeing a different future does not mean you can make a room see it. Asking a serious question does not mean the culture around you knows how to hold it. Sometimes the question itself becomes the problem.
The first problem was the objective. The next was the room, and that's where the next note begins.
Working through something and want a think partner?
· · · ·
What I'm claiming here
What I know: The letter in my mailbox, the PhD semester around it, my dissertation on the shareholder wealth effects of cause-related marketing, and the pattern of being flattened by someone with a little power, which I later recognized in myself, are all direct experience. So is the 2019 event study on AI acquisitions. I pulled the deals, saw the same handful of publicly traded companies showing up as repeat buyers, and found that many of the terms were sealed. That work was funded, presented at a conference with coauthors, and is cited in the source note below. My department chair’s comparison between AI and blockchain is something he said to me. The way I worked with these systems, building the context before asking and arguing with what came back, is how I actually worked. Corporate law does not require short-term profit maximization in every decision; see the source note. Stanford’s 2026 AI Index tracks rapid capability growth, adoption, and investment. McKinsey’s State of AI research continues to document a gap between adopting AI and realizing scaled organizational value.
What I think it means: A system organized around an objective nobody questions doesn't improve when you make it faster. The system just amplifies what it was already doing. AI can shorten parts of the path from description to prototype, plan, synthesis, or decision support, which makes the objective more consequential rather than less. The same capability, governed differently and aimed deliberately, may help raise the floor. The Human Readiness Gap is my working name for the distance between the power becoming available and our capacity to question its purpose, see and hear more of reality, imagine responsibly, and exercise judgment about what should happen next.
What I'm still unsure about: Whether the Human Readiness Gap is the right language for the pattern. Whether what I saw in the acquisition data was concentration of a kind that matters, or simply what any emerging field looks like when large firms move first. Whether the micro-macro tension I describe holds as broadly as I feel it does; the empirical literature on corporate responsibility, stakeholder outcomes, and shareholder value is genuinely mixed. Whether calling the AI future “branded” gives narrative too much causal power; my narrower claim is that stories shape what people fund, tolerate, demand, and attempt, which is different from saying stories choose history. And I am still learning how to build a mirror that can be challenged and corrected by the people represented in it.
Source note
Field Notes are essays, not academic papers — but when I make empirical claims, I want to keep receipts. My own background includes doctoral research on the shareholder wealth effects of cause-related marketing, which is where my interest in the Friedman question and the micro–macro tension began.
The AI acquisition work described here was funded by a Harold and Muriel Berkman Charitable Foundation grant through the Academy of Marketing Science and presented as Woodroof, P. J., Beachum, N. D., and Peasley, M. C. (2021), The Effect of Artificial Intelligence Firm Acquisitions on Firm Value, Dana E. Harrison, Editor. Society for Marketing Advances, 2021, pp. 125–126.
External background for this note includes Stanford University's 2026 AI Index and McKinsey's State of AI research on organizational adoption, governance, and value capture.
Public-company governance creates powerful duties and incentives around the corporation and its shareholders, but that is not the same as a universal legal command to maximize short-term shareholder return in every decision.¹ The leading US jurisdiction, Delaware, still treats stockholder welfare as the end directors serve, so this is a claim about time horizon and discretion, not about whom the law protects.²
¹ Paramount Communications, Inc. v. Time Inc., 571 A.2d 1140, 1150 (Del. 1989); Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682 (2014).
² Strine, L. E., Jr. (2015). The dangers of denial: The need for a clear-eyed understanding of the power and accountability structure established by the Delaware General Corporation Law. Wake Forest Law Review, 50, 761–789.
The interpretation through the Human Readiness Gap is mine and should be treated as a working Humanum Huis observation, not a settled research finding.
Comments