About
I made a bet in 2019.
It came due last year.
I am not a marketer who pivoted into AI. I chose marketing because of AI, and then spent seven years getting good at both.
The Bet
A guest lecturer, and one article.
In an undergraduate marketing class at the University of South Florida, a guest speaker walked us through an article about the artificial intelligence revolution. It reorganized how I thought about the next decade.
The reasoning I drew out of it was simple enough to bet a career on. Once AI makes software customizable by anyone, traditional software companies stop being defensible, because every business just builds what it needs. And when every product and service converges toward roughly equal, there is exactly one differentiator left standing.
Marketing.
So I stayed in it deliberately. Bachelor's and then master's, both in marketing, while learning AI on the side the entire time. A professor I was close with kept telling me to learn it as early as I possibly could. That turned out to be the most valuable advice anyone gave me in school.
The AI Revolution, Wait But Why. Still worth reading.
The Receipts
Early the entire time.
Fall 2022
Working in OpenAI's Playground and struck by what it could already do with copy.
30 Nov 2022
Day-one ChatGPT user. Have not stopped since.
2023
Early on Bard, which became my primary platform for a stretch. Then Perplexity.
Now
Claude, daily, as the system I build everything on.
Ask anyone who knows me. I was the person telling people where this was going about three years before they believed me.
The Adoption
Building it is half the job.
The other half is the part almost nobody writes about, and it is the reason my systems get used instead of admired.
It opens with a diagnostic question rather than a pitch. When I sat down with a reserve-study firm paying an agency a thousand dollars a month, I asked what the agency was actually doing for him, and two follow-up questions later he had diagnosed his own problem while answering me.
Then a working artifact before a proposal, an intake designed so a non-technical person succeeds at it on the first try, and a handoff meeting that ends with their own account configured and the thing proven in front of them.
A three-person operation now performs the function of a team of fifteen, and has already saved over a thousand dollars a month. We are still early on it.
The Turn
The market turned.
The thing that made me hard to hire now makes me useful.
I only ever worked at small companies, so I did everything. Email, social, web development, design, brand, positioning, project planning. For years that was a liability. Companies wanted a specialist, and a generalist read as unfocused.
The rooms varied as much as the work: a startup I co-founded, agency-side engagements, a hospital foundation answering to donors and boards, a private-equity-owned manufacturer answering to shareholders. I did not come at AI from technology. I came at it from watching how businesses actually operate, including the messy parts, from enough angles to recognize the patterns.
In the last eighteen months that flipped completely. Marketing teams are getting smaller and more capable because of AI, and the person who can hold the whole thing and build the tooling underneath it is now exactly what is wanted.
I did not change. The market did.
Hide In Plain Sight
The part that matters most to me.
Hide In Plain Sight supports homeless and at-risk students. I was the second student ever in the program, and the first to earn both a bachelor's and a master's through it. The program has awarded $700,000+ in scholarships, with a 92% success rate among its recipients.
I served on the board from 2020 to 2023, helping sharpen the brand, improve digital fundraising, and tell better stories about student impact. In 2023 they invited me back as the keynote speaker, which meant standing in front of a room and describing the distance between where I started and where I had got to.
It is the reason I think about opportunity the way I do, and why I would rather build someone a tool than do the work for them.
How I Work
A few things that are always true.
I do not do one-offs
Every request becomes a template, a skill, or a plugin, so the next one is faster. Solving something once and walking away is bad for the business and bad for me.
UNKNOWN beats a plausible guess
Every system I build refuses to invent a fact. A gap gets marked and surfaced. An accurate gap is worth more than a confident error.
I train rather than create dependency
The goal is that the team can run it without me. If a system only works when I am in the room, I built it wrong.
Rules become code
A style rule in a document is a suggestion. The same rule in the build pipeline is a guarantee. I move things from the first category to the second.
The Short Version