Technologist · Communicator · Educator
Twenty years building the systems.
Now asking what they cost to run.
I write and speak about the energy behind artificial intelligence — the power, water, and grid capacity that every model, every inference, every "just add AI" roadmap quietly assumes will be there. I come to this as a career technologist, not a spectator: software engineer, fractional CTO, and technical program leader who has spent two decades in the rooms where these tradeoffs actually get made.
Orofino, Idaho — relocating to Davis, California
Start Here
The research agenda, in five pieces
Where AI's energy cost actually lives: the product decisions that quietly set it, the business risk it creates, and the strategic case for treating it like the asset it already is. These are the first five arguments in that case — see all of it in Writing below.
Product & Business
The Hidden Line Item
Product and engineering teams make build-vs-buy and model-selection calls every week that carry a real energy cost — and almost nobody's pricing it in.
Contrarian Takes
The Efficiency Trap
Why efficiency gains alone won't shrink AI's total energy draw — an argument from Jevons Paradox, not from vibes.
Numbers & Explainers
Training vs. Inference, Actually Compared
Case Studies
Who's Actually Doing Grid-Aware Compute
Policy & Regulation
What Regulators Should Actually Do
Translating IEA and DOE data into concrete recommendations for utility regulators and city councils.
Writing
The full archive
Sorted by the question each piece is trying to answer, not by publish date. Check back for new pieces, or reach out via the contact form.
Product & Business
The Hidden Line Item
Product and engineering teams make build-vs-buy and model-selection calls every week that carry a real energy cost — and almost nobody's pricing it in.
Product & Business
Model Sizing Is a Budget Decision
Why choosing between a foundation model and a fine-tuned small model is an energy decision wearing a product-strategy costume.
Infrastructure
What a Data Center Actually Drinks
Power draw, water use, and PUE — the physical reality behind "the cloud."
Infrastructure
Carbon-Aware Scheduling, Plainly Explained
Policy & Regulation
What Regulators Should Actually Do
Translating IEA and DOE projections into concrete recommendations for ratepayers and utility commissions.
Contrarian Takes
The Efficiency Trap
Jevons Paradox says efficiency gains get spent, not banked. AI's "green" efficiency narrative needs to reckon with that — this is the argument for why.
Case Studies
Who's Actually Doing Grid-Aware Compute
A journalistic profile of a company treating power constraints as a design input, not an afterthought.
Career Pivot
What Twenty Years of Building Software Taught Me About Asking Better Questions
On being the translator in the room — technologist, teacher, and now, researcher.
Numbers & Explainers
Training vs. Inference, By the Numbers
Numbers & Explainers
AI's Share of Global Electricity, State by State
"The bill for not deciding always comes due. It's just late."— from "The Hidden Line Item"
Speaking
Topics I speak on
I'm building toward mid-tier sustainability-and-technology conferences over the next year. Local meetups, podcasts, and panels welcome now.
-
01
The Energy Line Item Nobody Puts on the Roadmap
A product-and-delivery framing of AI energy cost, for technical and business audiences alike.
-
02
Efficiency Won't Save Us — Here's the Math
A plain-language walkthrough of Jevons Paradox, aimed at people who've been told efficiency solves this.
-
03
Translating Between Engineers and Everyone Else
What twenty years of technical leadership teaches about making complexity legible without dumbing it down.
-
04
What Regulators and Ratepayers Should Actually Ask
A policy-facing talk translating IEA/DOE data into concrete questions for local decision-makers.
CV · Research Statement
Two decades, one throughline
Over my career, I've worked as a full-stack engineer, a fractional CTO for companies that needed technical leadership, and an IT instructor. As a technical project manager, I lead AI-enabled delivery programs, supporting 24/7 Follow the Sun team management and connection. Woven throughout my career, I've done start up, acquired, run, or turned around roughly thirteen businesses either by design, or for a stretch learning goal.
The throughline isn't any one title; I've spent two decades standing between technical reality and the people who have to make decisions about it — engineers, executives, boards, clients, students — and making sure the decision gets made with the clear picture in front of them, not a simplified or oversold one.
The one question I'm pursuing now is aimed at a big problem: what does artificial intelligence actually cost to power, and who's deciding without saying so.
Education
The Wharton School — Chief Technology Officer Program, Information Technology
University of Phoenix — B.S., Business Administration
PhD Target
Computer Science, with a research emphasis on data and energy — targeting UC Davis, tied to a confirmed relocation to Davis, California.
Skills
Technology & Platforms
Methods & Business Capabilities
About
The short version
I've spent twenty years as the person in the room who can talk to the engineers and the executives in the same meeting — sometimes in the same sentence. I've been a software engineer, a fractional CTO for companies that needed technical leadership before they could afford to hire it full-time, an IT instructor, and, for the last several years, a lead project manager running AI-enabled delivery programs at Tech9, supporting 24/7 Follow the Sun team management and support.
Alongside that, I've started, bought, run, or turned around roughly thirteen businesses — construction, PR and communications, home services, publishing, hospitality. Some of them worked. A couple of them didn't, and I learned as much from the ones that didn't as the ones that did.
What ties all of it together is translation: taking something technically real and making it legible to the people who have to decide what to do about it, without losing the truth of it in the process. That's the job I'm doing now, applied to a bigger question — the energy cost of the AI systems everyone is racing to build. I'm pursuing a PhD in Computer Science, with a research emphasis on data and energy, and building the public case for that work here, one piece at a time.
As for spare time, I try to focus my energy in community leadership and philanthropy, leading strategic planning for community organizations, building and leading small to large scale events and conferences, and find time for coaching and mentoring in the middle of all of that. Anything I can do to give a boost to the people who make our communities thrive.
Off the clock, I am happiest outside — hiking and cycling, floating the river, growing dahlias, and running my two border collies, Maple and Forest. I'm currently based in the woods of rural Northern Idaho.
Contact
Let's talk.
Editors, co-authors, conference organizers, and anyone thinking hard about what AI actually costs to run — reach out.
ccahorton@gmail.com