"...to get to a place where you could love anything you chose, not to need permission for desire, well now, that was freedom." —Toni Morrison I have an addictive personality. It’s just there, constant, whether you acknowledge it or not. Work, substances, people, screens, the next thing, and the next thing and the next thing. I’m wired to chase whatever lights up the reward center, and I’ve spent most of my adult life either feeding into that wiring or managing it. Usually, both at the same time. Pull the lever. Give it a prompt. Kick off a few agents. Push a build. Often, the output is garbage. Incoherent. Embarrassing. You adjust your prompt, pull again. And then, sometimes on the third pull, sometimes on the forty-second, it’s magic. Something decent and functional materializes that you did not fully make, but at the same time, could probably not have existed without you. Or that’s what you make yourself believe. The dopamine hits before you’ve even refreshed the output. After two decades of building software, you learn a thing or two about variable rewards. The feeling is the hook, and the highs are extraordinary. This month, I ran planning, research, design, frontend, backend, ops, and marketing across multiple projects. That’s not necessarily new. The difference is that I executed all those functions myself by delegating most work to Claude Code agents instead of direct reports. Three projects running in parallel, each supported by three to four AI agents working on different tasks across different codebases. When one wraps up, another has work ready for review. The pattern recognition, the resource allocation under pressure, the macro-micro switching all adds up. And it feels good. Not just productive-good. Physical. The screen at 2 am, the agents humming in terminals, the compression of an idea into something real with Massive Attack’s Mezzanine on the headphones. I love beautiful things too much, always have. And it’s never been easier to build really beautiful things. It feels like a superpower. Like something I shouldn’t enjoy this much. Very familiar. It also reminds me a lot of playing real-time strategy sims like Civilization or Age of Empires. Didn't think those wasted hours turned out to be Ender's Game-style management simulations for skill prep. But like every vice, there’s the comedown. What I learned before heading off to France to chase the toddlers for a week doesn’t matter the week upon returning. The workflows I thought I mastered in January feel quaint at best. Every skill I acquire feels ephemeral in a way that skills used to. I spent decades learning to lead teams. That knowledge compounded. Now I learn something on Monday, and by Friday, there’s a better way to do it that makes my Monday self look like he’s riding a bike with training wheels on. The work feels simultaneously like the most exciting thing I’ve done in a long time, and like I’m building sandcastles in La Rochelle at high tide. Shelf life of knowledge is shrinking to months, weeks, days. The same wiring that makes me chase things tells me this works out for people like me. People who adapt fast, who synthesize across disciplines, who were never just one thing. I see a clear future there, at least for a while. But I can’t look away from the fact that every function I performed across those projects this month used to be someone’s entire job. Dozens of those someones are people I’ve managed, mentored, and hired. Entry-level folks, Junior designers, junior developers already can’t get jobs. Right now. Today. The pipeline that fed knowledge work for generations is collapsing, and that’s just tech. Run that thread forward. Roles keep compressing, layoffs keep coming. Entire departments absorbed into one person operating a cluster of agents. No transitional infrastructure was built because nobody believed it would move this fast. Schools can't keep up pace. And the monetary system itself might need rethinking because the link between labor and income is fraying faster than anyone planned for. Meanwhile, a significant part of the population is spending their collective energy arguing about bathrooms or in armed conflict.
It’s been said many times before, and I want to be part of the optimist camp, but I think we are deeply, dangerously underestimating how fast this is moving. Not “we” as in people in tech. Everyone. The mismatch between what’s happening and our collective ability to process it might be the defining crisis of the next decade. You’ve read about this as the singularity, I’m sure. Just like everything else, it won’t be evenly distributed. I’ve stayed grounded in February by knowing I teach my sons things that won’t change. I teach them to create, to enjoy. To love, to care. I teach them to draw. To sing. I teach them how to ride their bikes and how to ask for help. Respect for others, for themselves. These aren’t skills that expire. I’m confident in this. So I'll rely on them to get me through March. The promise of AI is abundance. The system we’re all running on was designed for scarcity. What happens when scarcity dissolves? I know billions of people need clean water before they can think about abundance. That gap has to close first, and I believe these same tools can help close it if we choose to use them there. But for those of us lucky enough to be even having this conversation: stop pretending that one of anything was ever going to be enough. One skill set, one identity, one way of being useful. The people who do well in what’s coming will be the ones who permitted themselves to want more than one thing at a time. Maybe that’s naive. I’d rather be naive and build than pretend none of this is happening. If you’re building too, let’s talk. Things on the internet I found interesting lately:
|