Ray Bolouri.

I started programming at eleven, on a computer that plugged into a television. The first real program I wrote was a stripped-down word processor, in assembly, from a book. Then I photocopied more than a hundred one-page resumes looking for a free internship, and heard back from no one. I kept building anyway. That is the pattern the rest of my career runs on: build the thing first, and let the world arrive later.

I have spent forty years building the layer underneath the visible one - the routing, the protocols, the wireless links, the texting rails, and now the voice and agentic systems. People told me a fast enough parallel machine was not worth building, that wireless would not reach that far, that texting a business was a stupid idea. Each time, “impossible” turned out to mean “not yet engineered.” So I stopped treating it as a verdict and started treating it as a to-do list. The useful question was never can this be done - it was how.

That is why I read technology from the plumbing up. If you have run the cables and watched the latency, you stop being impressed by demos and start asking what the system actually routes, and where it breaks. It is also why I am steady about AI: machines are extraordinary at repetition, recall, and precision; people are extraordinary at judgment, creativity, and care. The good work is putting each on what it is built for.

I also had to learn the things school does not teach a builder. I was the shyest kid I knew; I learned social skills on purpose, the way I learned Pascal. I taught myself finance, how trust is built, how to stand up for the people who backed me and, eventually, for myself. The lesson underneath all of it is the same one: hard things are learnable. That is the whole optimism, and it is the reason I think the next era is buildable too.