AI Made Simple

I did not mean to become an AI guy. I want to be clear about that upfront.

It started because my manager used the phrase "leveraging AI capabilities" in a meeting and I nodded like I knew what that meant. Then I had to actually know what it meant. So I bought a book.

The book is called AI Made Simple. I read it on the subway. I held it up in front of my face the way you do when you are absorbing information very seriously. People around me were staring. I assumed it was because I looked like a man on the cutting edge of technology.

It was a good book. Very clear. By the time I got to my stop I had learned that AI is a "transformative paradigm" that will "reshape the nature of work itself." I did not learn what a transformer was. The book assumed I would look that up separately.

What I Did Next

I went home and opened ChatGPT. I typed: "explain AI to me simply."

It wrote four paragraphs. I read them. I understood all the words. I did not understand the thing. I asked it to explain AI even more simply. It wrote a paragraph about how AI is "like a very smart calculator that learned from examples." I thought about this for a while. It did not help.

I asked Marcus. Marcus said AI is "gradient descent all the way down" and then left the room. I am still thinking about that one.

The Framework

The book had a framework. Most business books have frameworks. This one was a four-quadrant grid with axes labeled "AI Maturity" and "Value Delivery." You were supposed to plot your organization on the grid and then move toward the upper right quadrant where it said "AI-Enabled Enterprise."

I plotted my organization. We were in the lower left. The book called this the "AI Curious" stage, which felt generous. A more accurate label would have been "has heard of ChatGPT and used it to write one email."

The path to the upper right involved something called an "AI Center of Excellence." I showed this to my manager. He asked if that was something we needed to buy. I said I thought it was something you built. He asked how long it would take. I said the book suggested eighteen months. He said we would revisit in Q3.

It is now Q2. We have not revisited.

The Prompts

The book had a chapter on prompting. This is apparently a skill now. You have to talk to the AI correctly or it gives you bad answers. This felt like being told you have to pet the calculator in the right direction before doing math.

There were twelve prompt frameworks in the chapter. I wrote them all down in a notebook. I used none of them. I have instead been typing things like "make this email sound more professional" and "can you make this shorter" and "no shorter than that." The AI complies. It does not seem to care about frameworks.

My best prompt so far was "write a status update that sounds like I did more than I did." The AI wrote an excellent status update. I sent it. Nobody noticed. I am still thinking about what this means.

What I Actually Built

The book said I should "build an AI prototype to demonstrate value." I built a thing. I am not sure what to call it. It is a Python script that takes a spreadsheet, sends each row to an AI, and asks the AI to summarize the row. The AI summarizes the rows. The summaries are shorter than the rows but contain the same information. I have not demonstrated value. I have demonstrated that the loop works.

I showed it to my colleague Janet. She asked what the summaries were for. I said they were for reading. She said the rows were also for reading. I said yes but the summaries were faster. She said she could read the rows pretty fast already. I said yes but imagine if you couldn't.

Janet did not fund my AI Center of Excellence.

Where I Am Now

I am on the subway again. I am reading a different book. It is called AI Strategy for Leaders and it has a different four-quadrant grid. One of the axes is "Strategic Alignment." I am going to learn what that means.

The man across from me is reading something on his phone. He keeps glancing at my book. I am holding it up high so he can see the cover. I want him to know I am taking this seriously.

I think I am in the upper left quadrant now. The book calls this "AI Ambitious." Apparently I need better data infrastructure before I can move right.

I will ask my manager in Q3.