I Cloned Myself. It Hasn’t Made a Dime.(How I AI 3 of 4, Andrea’s Version) (Copy)

About AJC

Nov 11

How I AI

Andrea’s version · a four-part series

Part 3 of 4 · November 2026

About 2 years ago, I built an AI version of myself, and so far it has made exactly zero dollars. Here’s why I did it, what I expected, and whether I’d do it again.

(This is Part 3 of the series. Part 1 covered the tools I use; Part 2 covered what I’m building.)

Why I cloned myself

Brian Halligan, the co-founder of HubSpot, put an AI clone of himself on a platform called Delphi. It’s trained on everything he’s written and said, and anyone can ask it questions. I watched that and thought: if it works for him, it might be a good idea for me. Plus, I loved the idea of "locking up" my name on Delphi in case it became a big deal (more on that with andreajones.ai).

So I built one. I trained my Delphi Andrea clone on everything I've ever written, published, or video-ed about Executagility, ERP selection and implementation, project execution, and change management. It scrapes my LinkedIn, my websites, and my Youtube sites. At first, I kept a Word file with links to everything I uploaded to my Delphi "Mind," but then I stopped because I can literally see it all listed in the back end of Delphi as well.

What I thought it would do

Answer common questions at 11pm. Warm up prospects before a first call. Maybe become a paid product for people who want my frameworks on demand.

What actually happened

No one in my “ICP” finds my Delphi at all unless I direct them there during a speaking engagement or when I link to it on a LinkedIn post. Some random person found it and asked for advice on how to apply to MIT’s LGO program, and someone else asked it IN SPANISH how to prioritize projects; turns out my clone is fluent in Spanish! LOL.

What the market is telling me

There are few things. First, Delphi is a startup, and definitely not mainstream. I met an entrepreneur here in Orlando trying to do the exact same thing on her own (AI clone for coaches, thought leaders, consultants, and coaches) and she hadn't even heard of Delphi. (Don't worry, I shared as she needed the info for her competitive analysis!).

Second, the idea that you can get coached by an "Expert" might be appealing for people like Brian Halligan who truly has done some amazing things, but I'm not that famous. I can barely get people I know to take calls with the real me, why the heck would they want to talk to a robot version of me? Ha!

My Claude generously offered that maybe it is how people are thinking about AI right now. As in, people are a little turned off by the quasi-human nature and if it's going to take over our lives. I think that's a great testament to the sycophantic nature of AI. Gee, thanks, Claude, for trying to make me feel better about myself. It's okay, though, I can take the cold hard truth, my Delphi is a novelty and there is no ROI.

Still - here are a couple of stats that you might read if you don't think AI found them: (Spoiler alert, it did! But I wrote the above paragraph)

The trust problem, in two numbers

69%

of people who consume online content say they trust AI-generated content less than content made by a person.

Source: Pangram Labs / YouGov, April 2026

61%

say they’re unlikely to read or engage with content at all if they think AI made it.

Source: Pangram Labs / YouGov, April 2026

Which raises the obvious question. Why would a CEO want advice from a simulation of me when the real one is a phone call away? For a celebrity with a million followers, an AI clone might be a business. For a consultant whose value is judgment in the room, I’m not convinced it is. Too bad I'm not an Influencer!

Where I’ve landed, for now

The clone is a possession with potential, not a revenue-generating-product. I'll keep it as long as I think there is value in owning it, and not longer. I paid for the annual subscription, so it's there at least until May 22, 2027. And I did add it to our Contact page for AJC here, as well as the Executagility site under AI Trainer here.

And yes, I bought andreajones.ai. That one I’m keeping indefinitely. It costs very little, and it is my name. Right now it points to the AJC website, though I am debating whether it makes sense to pay to host it on its own.

The 90-day test I now use on every AI subscription

1. Does it do something that NOTHING else I own already does (or will do in the next 90 days)?

BTW Claude loved this one when I asked for feedback saying: "Question 1 is the best of the three and the most original. Most subscription advice asks “am I using it?” Yours asks “is this about to become a feature of something I already pay for?” That’s exactly the right question right now, because so much gets absorbed into ChatGPT and Claude within months."

2. Do other people monetize it, and does it make sense that I could too?

3. Will it "spark joy" in a way that would not be sparked without paying for it (and not break the bank)?

I don't really want to keep piling on subscriptions, especially as I think the general models I already own through the "big players" of OpenAI and Anthropic are going to continue to add functionality. In that way, they are "sticky." However, I do wonder if they will drive up costs for individual consumers or small businesses in a way that I won't be able to afford, such that enterprise aggregators renting out specific functionality built on their platforms will be the way little guys like me get to use their software in the future. I'll cross that bridge when I come to it.

Maybe I will ask Chat and Claude what they think.