The New Frontier – Why I Bet on AI Before It Was Cool
Somewhere between November 1994 and January 1995, I got my hands on the Netscape browser. Every click on a hyperlinked, static web page sent my imagination further than the last. I don’t usually bring the technical details of my job home, but that day was the exception.
‘This is going to revolutionize the world,’ I announced the moment I walked in. My wife assumed I’d taken something stronger than tea.
Fast-forward to January 2018. I finished an online AI course and made the same declaration with the same grin. This time my wife just said ‘hmm.’ That’s the tell worth paying attention to. The second time you recognize a genuine platform shift, you recognize it faster and with less noise around it. The pattern itself becomes a signal.
Here’s the uncomfortable part for anyone chasing an ‘AI-first’ narrative: the underlying concepts, architecture, and design of artificial intelligence have been around for more than seventy years. I’m not that old, just the ideas are. Back in the mid-1980s, on UNIX, I built a crude ELIZA/DOCTOR clone in C and installed it as a boot-time prank on our office machines. It fooled my coworkers for a good while, right up until it started looping the same reply, not unlike Siri when it can’t place your accent.
What changed isn’t the core idea, it’s the economics around it. Big data, cheap and abundant compute, and mature tooling created the perfect storm for consumer level AI. What used to require laboriously hand-built datasets, begged-for compute time, and every function written from scratch now ships as a purchased dataset, an assembled GitHub repo, and as many VMs as you’re willing to pay for. Call that the Inflection Test. A technology is genuinely at an inflection point not when it becomes technically possible, but when it becomes economically trivial for anyone to attempt it. AI crossed that line; most of the last decade’s hype cycles didn’t.
After more than thirty years in hi-tech, I stepped back to do something non-technical for a while. But a techie doesn’t get to let ideas simmer indefinitely, so this blog, ‘AI – The Next Frontier,’ is where I put them down. Read on, pass it on, build something from it, or push back on it. If you’re trying to work out which parts of your organization’s AI enthusiasm are signal and which are static, that’s exactly the conversation I’d enjoy having.