Still Early

I head Data Science and AI for a startup in a "traditional" industry that lags the SOTA by about 5-10 years in terms of tech adoption.

Demoing the analytics portions of the product sometimes gives me unique insights into what "regular" non-AI folks - plant managers and operators - care about.

For instance, today I demoed the same new feature we're hoping to get to market w/ 3 UXes:
1. Static dashboard with KPIs and metrics
2. Static deep dive dashboard with KPIs, charts, stories and
3. Interactive chat-based dashboard w/ dynamically generated content (graphs, text, KPI cards) via LLM interface to the data.

When I asked our client for feedback on which flow he preferred, he said very matter-of-factly: "None - I like spreadsheets! I haven't used AI much and don't quite get it. I've used spreadsheets all my life and it's worked out just fine."

Now this isn't representative of all the folks I've spoken to, but each such interaction grounds me w.r.t what people actually need versus what can be built right now.

We're still in the very early innings of an AI adoption process that's going to play out over another generation in multiple fields.

If you're starting out, you might want to stop before you FOMO into doing AI things that you feel like you must because everybody else is.

Same lessons for crypto. We are very early in both trends. True adoption will take a generation at the least. Less for AI, more for crypto in my estimation.

On the way, you'll find many opportunities to contribute your talents and earn a good living. Breathe deeply and relax. It's going to be fine.