Part 1
Why I started paying attention again
The ground is shifting under our feet again. Every big change in technology works this way, and the people that tend to feel it most are the ones accidentally standing in the wrong place. I was in telecom when dedicated trunks gave way to IP switching. I was at Salesforce when software stopped being something you installed. Some of you were there with me, everyone has been impacted by it.
I was a functional solutions guy. Before the “architect” title came into play. When the work was about figuring out what the system ought to do and how people interact with it, and then handing the building to people better at it (much better at). I am not an engineer, a coder, a person that can follow A-Z step-by-step without veering and I have never claimed to be.
Then I spent eighteen years in mineral management, running Lodemark, in a corner of the world where the software, quite frankly, is not very good. I built a management platform for the firm because what I could buy did not do the job, and I have been tinkering with it for years, because the design part, where you solve problems, is the part I like most.
Over those years, I dug into most of the technologies that came and went. Some were big, like blockchain, but I never saw how they changed anyone’s day-to-day work. Others were flashes in the pan. None of them really grabbed hold. I was raising kids, coaching soccer, running a business, and everything else that goes along with trying to be present in your own life. But the designer in me kept an ear up, kept tinkering.
I’d had a vision for something once, but never the means or talent to build it. Then, like everyone else when the first consumer versions of AI came out, I played around: a soccer team logo, a recipe, a business problem. Clearly, something with this new wave was different, but I couldn’t put my finger on what I could do about it. So, I added technology conferences back to the calendar with the industry ones. Then a conversation with a good friend, and consultant worth your time, Rich Nazzaro, sparked my curiosity, my need to mess with things again. We talked about what was coming, and the parallel to the moment SaaS took hold.
It really wasn’t the technology that got my attention about language models or “AI”. It was that some of the same questions that followed SaaS or other technological shifts around for years were rearing their heads again, just with different syntax. The biggest one, as always:
How do you know you can trust what the thing tells you or what it’s doing?
That question is not academic where I sit, and not unique to my position. Our job is to hand a client an answer about their own property, their own resources, assets, and money, and provide them with the best options so they can act on it.
Since January, I have been diving into it twelve and fourteen hours a day. Not to show how hard I work, but because I am self-taught (always have been) and the only method I know is to find what breaks and what I can fix. Or in this case, find out what breaks, what works, or seems to work, only to find out it’s fool’s gold and fix it again.
The models got better as I was working with them and we all know they are moving faster than we can keep up. Some of what I ran into has since been closed by the models themselves. But, I know why those problems existed, not just what the workaround was.
Plenty of what I hit is obvious to somebody more technical, and that is fine, that’s not what this is about.
Next week we’ll look at what I really f’d up. Wonder if any of you share the experience, or what was just my own dumb mistake thinking I’m smarter than I am.