From Error to Virtue

How to be creative

This was a fun week - I spent it in Auckland, a really lovely city full of really wonderful people. I was here to give a talk and help with a hackathon for a company here. The topic, of course, was AI, in many forms.

The hackathon was interesting. It used to be that you had to have very strong, ready technical skills to build something meaningful quickly (I once counted the number of programming languages I’ve learned and stopped when I got to 10). This time, I was able to do quick bits of creating while the teams were explaining their ideas at the judging portion at the end! (Nothing too amazing - for one team I expanded on their idea a bit and came up with a number of better names, and for another I wrote a theme song and generated the music).

I believe fairly strongly that one of the core skills of the coming era isn’t going to be something like memorizing a programming language, or being able to follow complex instructions, or fit into the status quo…it’s going to be the ability to imagine, create, and have initiative and agency with problem solving and tools.

In the course of a conversation here, I was talking about how we all have “infinitely patient, world-class experts” in our pockets now, and saying that there is less and less excuse for not building something. Don’t know how to code? That’s fine, the models do. Don’t know even what code is or how it works, or how to set it up, or what a language is, or which one you want…the model will answer all of those questions, as long as you ask them. Maybe code isn’t even the right thing - that’s fine too! Tell it what you want to build and follow the breadcrumbs, just being curious.

As I was saying all of this, the person I was talking to suddenly had an “aha!” moment - that all of the ideas they’d been having over the years could now be built! They didn’t have to ask a programmer or convince anyone to do a demo - they could just do it themselves. It was amazing to see this happen.

And this is the skill to develop - some combination of imagination and persistence, and active sense of making things rather than a passive one of waiting for instructions or “staying in your lane”. In my own making, always, I have a sense of impatience and experimentation. It doesn’t always work out - hence the title of this post, “make virtue from error”. It’s not the success of the next step that matters, but that you are taking it, and the one after that, and that each time, you learn something from what doesn’t work for the next attempt.

I could see this in the hackathon - some teams (and individuals) had that sense, but many didn’t. To most, it was an afternoon exercise they were being asked to do, which they would do well enough to make some modest progress but not with real energy. Understandable, not everyone wants to create new things all the time - it’s hard!

But I can see within the teams of engineers I work with on a daily basis: the folks who come to the table ready to try, ask, experiment, imagine and iterate are starting to go much faster than the folks who are passive, getting-by, status quo, using familiar tools. The tools are starting to feed back on themselves, too - good engineers are starting to use AI-based tools to make their AI-based tools even better, and so on. It’s hard to know where this will end but it will be very powerful. When you find a rocket, get on it!

There are still lots of challenges and problems with AI. It will be years, probably before LLMs are as fully mature as, say, the smartphone is. But that’s fine - that’s how tech works! Learn now while it’s moving and evolving - it will serve you well as it matures and grows in effectiveness. The skills of the future aren’t memorizing, they’re imagining, making, iterating, and learning.