Cruising the YouTube highway.
No offramp in sight.
People keep asking if I’ve watched Love Story, the JFK Jr. / CBK mini-series that has made a few NYC spots certifiably unbearable.
I have not.
I’ve been watching a lot of YouTube.
I am (well, I had been, until last week) on cruise control on the YouTube highway. I’d been watching all kinds of signs and billboards go by, waiting to see my exit, but it didn’t come — and I’m not sure it would have, tbh.
YouTube has good stories, you guys!
Chris Koerner on The Koerner Office, Starter Story, … these are fairy tales (loosely speaking, don’t @ me) for the self-motivated.
I was learning about how others spotted their use cases, how they applied AI and automation to different patterns and pain points they spotted in their own lives. I kept cruising from one video to the next, thinking eventually I would have a moment where a pattern for my own life would pop into my head.
That didn’t happen.
As I said in my last post earlier this week, YouTube gave me a destination with no map. At best, it’s a highlight reel from other people’s maps.
Conceptual knowledge (~knowing what) is not mechanical knowledge (~knowing how).
In medicine and probably elsewhere, there’s the notion of see one, do one, teach one. The same thing applies when it comes to finding relatively high-value use cases for AI in our personal and/or professional lives (this is incredibly unsurprising to everyone who has done one and taught one); it’s just that making the leap from seeing one to doing one is harder than it seems.
The value in making said leap became exponentially clearer to me in real-time last week as I started to chip away at the foundation for the product mentioned in this post: pattern matching to personal and professional use cases based on conversational diagnostics aka leap enablement, if you will.
I think this—caught between seeing and doing one—is the canyon that most knowledge workers live in. Most people I talk to have AI conceptual knowledge. Most have used Copilot to summarize documents. But that’s like limiting your computer use to using solely a word processor.
Most people I talk to don’t have AI mechanical knowledge. Many are angry about slop and being a janitor of poorly formed prose. They’ve yet to cross the canyon from understanding to application. Why? Because it’s hard and abstract. It’s knowing you should work out without knowing which exercises to do.
This is why programs like Hilary Gridley's Couch-to-5K for AI work. They put structure on the abstractness via small daily reps anchored to mechanics rather than concepts.
Like virtually everything, it takes my (your) own clicks and trips to truly understand the tips and tricks being offered by others.