https://x.com/jasich/status/2115753596524560644
I started my little vacation last week with this X post from my colleague Jason, who is objectively great at that AI Engineering thing that we have going on. I've been thinking every day about this post since, reflecting and correcting.
If we look at the past couple of years, we've been seeing a bunch of tools and techniques coming up to "bridge the gap" to better quality or faster software. Think of the plan mode that served us so well at the beginning of Claude Coding. That bridge is mostly burned. Most devs don't use plans anymore (in my bubble) because the agents are smart enough to figure it out. Or, think of worktrees: cloud environments are coming for them, sooner or later. If you think of some other random piece of new technology like Jev, it was hot for a minute. Then, right after the public launch, every big provider caught up with their own decision model. Same for the "software factory". There will be a solution from all the major token labs to make it more plug-and-play. With all those engineering pieces, you can get a bit of an edge for the moment, but the how-to-engineer stuff gets old quickly anyway. Everything will be upside down in a few months, again.
So, if you want to build something lasting, that's a different game. You gotta be more "ambitious", they say. I like the idea, but I have to admit that the projects I've taken on in the last year haven't made me uncomfortable enough. If you like the idea as well, have you asked yourself why it's so hard to be more ambitious in a time when nothing seems to be out of reach? Ultimately, engineers want to build something useful or entertaining. Now, anyone could, but I'm not sure that value or entertainment has increased proportionally to the increase in building power in the last 1-2 years. Doing the "ambitious" thing and getting uncomfortable is ultimately something that is hard to do: Being uncomfortable. Thinking big. Persistence and discipline for the pursuit of quality. Plus, the fundamentals of building something of value or entertainment haven't changed. It's still potentially a years-long, for many, lifelong effort to find that something that clicks.
Still, I would love to explore ambition and discomfort a bit more in the next newsletter editions here, while keeping on sharing AI engineering nuggets around scaling my outputs at ClickFunnels when they come up (i.e., "software factory" or whatever we're gonna call it by the end of the year). At this point, I'm also maintaining a few in-use production apps, a few open-source libraries and a few local apps I use daily, but they seem too unambitious and too niche to be worth sharing right now. We'll also tone down the link sharing here. Those seem just for fun these days anyway. You will probably agree that we need less information and more focus.
I'm looking forward to what we are going to build here with more ambition and discomfort in our backpacks. I wish you a good focus.