It's 1995 All Over Again — by Dan Lynn
By Dan Lynn
In 1995, you could feel something enormous beginning. The internet was just arriving, and the possibilities gave me goosebumps. I didn't have it all figured out — nobody did — but it was impossible not to sense that the world was about to change. And it did, more than any of us imagined.
I've waited a long time to feel that again. Now I do. AI is that moment — the same crackle in the air, the same sense that everything is about to be rebuilt, the same goosebumps. It's 1995 all over again.
What I took from the first time around is simple: when a moment like this comes, you don't wait for certainty. You lean in, start building, and learn as it unfolds. None of us can predict where it goes — but it's a privilege to be here for it, and worth making the most of.
Questions this Key Learning answers
How do I start building with AI when the future is unclear?
The author’s lesson from the early internet is not to wait for certainty. Lean in, start building, and learn as events unfold. Our analysis suggests beginning with a real customer problem that generates fast, usable feedback. Early builders gain judgment from customers, failures, and constraints—information that people watching from the sidelines never receive.
Should I use AI to automate work or redesign it completely?
Our analysis argues that AI’s deeper value will come from rebuilding products and workflows, not merely adding a chatbot to existing processes. The author sees this as a moment when everything may be rebuilt. The practical takeaway is to treat AI as a reason to redesign work, rather than only automating the steps already in place.
How should I judge an early AI project with little revenue?
Our analysis says early efforts can be judged partly by what they teach, not only by immediate revenue. The first move does not need to reveal the final destination. As the author puts it, the point is to start building and learn as the technology unfolds, developing better judgment through real customers, failures, and constraints.