2026 AI Journey So Far
Three "Aha" Moments in How I Work With AI
AI snuck up and tackled me in January.
In 2024 and 2025 I’d experimented with chatbots a bit, and was not overly impressed. Saw some potential but much of what it produced was, well mundane.
OpenClaw and Claude 4.6 started getting some attention and I took another look - wow. We are not in Kansas anymore!
Since then I’ve been going deep. Trying every tool, learning techniques, building internal tools to really get a handle on what these things can do. Now I’m ready to start building real things — and I’m going to share it all here as I learn.
Aha moment #1 - You don’t need all the answers
In most previous jobs I was expected to have the answer. Do research if needed, solve the problem, make the call.
The first “OMG” moment I had with AI was when I gave it some really vague (honestly, crappy) directions and told it to just take a swing.
It did. It was not everything I would have done but it made a lot of really good choices and got the thing to 80% on its first attempt. It was WAY easier to iterate and edit rather than staring at a blank canvas.
Aha moment #2 - Flip the script
Not my original invention but try something like this: “I want to start a business but I’m not sure what to build. Here are some areas of interest [list] and here are some things I’m pretty good at [list]. ASK ME QUESTIONS to help me get unblocked and figure out what kind of business is best for me.”
Asked me maybe 8 (pretty good) questions and came up with a ranked list of things to explore, which we then did over the next week.
You can have the AI actively guide you through some very open-ended problems.
Aha moment #3 - Help me help you
I’ll expand on this greatly in the next post, but let your AI give you feedback as well. At one point I asked it “hey is there a better way to handle this” and it did two things, one of which was very surprising. First, as I expected, it went over a few aspects of what we were working on and provided some key points about decisions that were made and then offered a few suggestions.
Then it went “meta” on me and gave me this:
As Claude put it: “Indecision is the slowest tax on AI-assisted work.”
We are not in Kansas anymore.
Steve Reaser helps small businesses put AI to work. He writes about what’s actually working, and what isn’t. Watch me learn and build in public at SteveReaser.com