What they said would happen

In August 2022 I wrote down three things people had told me I would not finish.

Going to the gym to reach a certain physique — “probably you won’t make it, you’ll quit in a month or two.” Time passed, goal achieved.

Starting to learn to code — “it’s too difficult, you’ll get overwhelmed and quit soon.” Time passed, goals achieved, several languages, many projects.

Building a tech startup from scratch — “probably you won’t make it, you’ll get exhausted and quit soon.” Work in progress, let’s see.

I left the third one open because in 2022 I genuinely did not know. It would have been easy to write something confident. I did not have the outcome, so I wrote “let’s see.”

It has been four years. I can close it now.

Closing the third one

I did not quit.

What I will not do is convert that into a victory lap, because the honest version is more complicated and more useful. The startup I was building in 2022 is not the thing I do now. What actually happened between then and now was a consulting role at a legal firm, then a stretch as an AI consultant, then two years as CTO of a technology company, and now running my own again.

That is not the arc anyone plans. It has more turns in it than “built a startup” suggests, and at several points it would have been reasonable to describe it as not working.

So the accurate answer to the 2022 prediction is: they were wrong about the quitting and roughly right that it would be harder than I expected. Both things are true. The prediction that I would get exhausted was not stupid — I did.

What they got wrong was assuming exhaustion is the end of the story rather than a recurring feature of it.

Why I keep the receipts

I had written a roadmap a few months before that post:

• 2020 — learn programming from scratch [Done]
• 2021 — learn mobile development and build the app [Done]
• 2022 — build the startup and grow it [In Progress]

Publishing that was uncomfortable at the time, and it is the single most useful thing I did. Not because it held me accountable in any dramatic way, but because it makes this post possible. I can point at what I said I would do, and at what happened, and the gap between them is checkable by anyone who wants to check it.

That is also why I put numbers in everything else I write. When I say a vendor quoted $1,000 a month and we delivered the same capability for about three cents a page, that is not a motivational claim — it is an invoice. When I say AI took a rebuild from two months to under thirty minutes, that is a timing I recorded at the time.

Motivational writing without receipts is worth very little. I have read a lot of it. The reason I am comfortable writing the encouraging posts is that I can show the working behind them.

The two other predictions

There is a version of this pattern that is not about strangers on the internet, and it is the more damaging one — the prediction made by someone with authority.

In 2019 I had an idea for letting patients get prescriptions approved before reaching the pharmacy, instead of queuing for half an hour while the insurer responded. I described it to a marketing manager at a large pharmacy chain at a job fair. He told me it was very complicated and would never be implemented.

In November 2021 I walked past a pharmacy and saw a poster advertising exactly that.

Not by me. I want to be clear about that, because the story is not “I was robbed.” I never built it. The point is narrower and worse: someone with more seniority than me was confidently wrong, and I let that confidence stand in for analysis for two years.

The second one is smaller and stranger. I once sat two interviews with substantially the same CV. In the first, the interviewer understood the work and told me I was overqualified. In the second, the interviewer did not, and I was marked unqualified.

Same person. Same skills. Same week. The variable was not me.

What I would actually say to someone

Not “ignore everyone.” Some of the people who told me things would be hard were right, and a few of the warnings I dismissed cost me time I could have saved by listening.

The rule I would give is narrower: treat a prediction about you as a hypothesis, and notice whether the person making it has any evidence. A senior title is not evidence. A confident tone is not evidence. Having actually tried the thing is evidence.

And then go and generate some evidence of your own, because that is the only input to this that you control. It does not have to be impressive. My first piece of evidence was a script that edited my cover letter. Four years later it was a roadmap with two items marked Done and one left open.

Write down the open one. Come back to it later.

You would be surprised how often you get to close it.