Notes – Finishing Things

A lot has been happening in my life, especially when it comes to my career. I didn't have that much time to write (or so I tell myself), but there's one realization I really want to tell you about.

Since I'm semi-responsible for artificial intelligence in two different companies, I consume a lot of information about it. Just like in the character of the modern world, there seem to be two camps – and both of them are on the extremes. Makes sense: grand statements create clicks, which then in turn create more usage for these AI models. One week we're all going to be out of a job by Thursday, then Thursday rolls around and I'm still fighting with an Excel-sheet someone created twenty years ago. The trough of disillusionment (or at least part of it) of the Gartner hype cycle is upon is. One thing can't be disputed, though: AI has given us much more capabilities than we had before.

Just like when the computer became mainstream, it changed what kind of skills the workforce needed to succeed in the market. One of the main discussion points is the skills the workforce will need in the future because of AI. It's not clear what kind of skills AI may be able to do better than humans, but one thing is clear: these models don't follow their own agenda. No, they're not conscious, so they do what people tell them to do. So, one logic goes, since these models may be able to perform tasks better than humans in the future but still need guidance, the greatest skill one can have is agency.

But in my opinion, this leaves out at least 90% of the equation. Having agency is not enough when agency means only starting.

There seems to be two types of agencies:

• Sprint agency
• Marathon agency

and we need both to succeed.

Sprint agency is the thing that gets you going on a project or task. Just like fast twitch muscle fibers, sprint agency can move heavy mental weight quickly, but only for a short time. It punches what Steven Pressfield calls "The Resistance" (a force hindering you to start endeavors) straight in the face. But that's nowhere near enough. Starting may be one of the hardest parts, but it's not what will get you where you want to go. You might even make the argument that starting has also been made easier thanks to AI. You don't even have to do your own research anymore and acquiring skills is now easier. "An object in motion stays in motion" might be true in a vacuum, but engaging in a project does not happen in empty space. Resistance is a tough motherfunner. After you've Mike Tyson left-hooked it in the face, it gets right back up.

That's when marathon agency comes in.

Marathon agency is the ability to keep pushing through even though the shiny effect of something new has worn off. It's the follow-through with the endeavors you start. Starting to create an app is not the same as getting it to production. My grandfather always told me there are two types of punctuality: on time or late. Nothing in between. It's the same with projects. A project finished up to 80% is in the same state as one finished only to 1%: unfinished.

Of course there are is some nuance to this. You will still take the learnings you made during a project, even though you might not finish it. Also, knowing when to quit something is a skill in itself. Finishing something isn't always the winning strategy, otherwise the sunken-cost fallacy wouldn't be a thing. But this philosophy can also become an escape hatchet into the morass of might-finish-later.

I know how this feels: I have a graveyard of unfinished blog posts. But on average, finishing these ideas is not as fun as starting something new. There might more excitement in starting, but seeing ideas through to the end is more satisfying. I believe this satisfaction comes from overcoming struggle. The brain pulls this cute trick where it presents the new idea as smooth-sailing from beginning to end. It never is. Even if you start the most exciting new project in a domain you love, facing roadblocks is a given. Yet exactly these roadblocks are the seeds for satisfaction to blossom. Watering these seeds means overcoming these challenges.

Agency is the roots, but finishing things lets the tree grow. The more trees populate a forest, the more seeds will be spread. So the more projects you will finish, the easier it will become to not only become more agentic, but also strengthen the resolve to see things through. Whatever might get automated, I believe there's something AI can never replace: the human will to create things – even when this means struggling from start to finish.