First a confession.
Until about March of this year (2026 as of this writing), I thought that I was in a race against AI to make as much of a nest egg as possible until all demand for human thought evaporated. In that race, I thought that it was important for me to leverage AI as a work multiplier, to help pull ahead as much as possible. Apart from the selfish “I got mine” attitude inherent in such an approach, which was, at its core, driven by fear of irrelevance, I also found that, at least in my daily life, my cognitive skills were getting duller.
I was years removed from having reading (and I don’t mean incidental reading as a byproduct of scrolling) and writing as a daily practice. I noticed my vocabulary shrinking. And my go-to before-bed activity had become doom-scrolling Reddit, feeling mostly unsatisfied with every vertical swipe of a finger.
At one point, I had to stop and ask myself, “What the hell are you doing, Jesse?”
I’ve been using AI since the days of GPT-2, when a startup had latched onto the potential for using the LLM to mass-produce text for the purpose of ranking blog articles. And in using it, I felt both dirty and exhilirated. Here was a magic autocomplete that cut my time to produce marketing emails by maybe 90%. Maybe the emails weren’t great, but my goodness was it easy! I didn’t have to think as hard, and I could turn my attention to other, more important things. Although at the moment, I can’t remember what those more important things ended up being. Probably I ended up scrolling more on Instagram in my downtime; I doubt I ended up spending more time on my bicycle with the wind in my face, or standing next to a lake with a fishing rod in my hand.
AI got faster and “better.” It became a bullshit generator that could not only write passable bullshit, but it also could write code that would run. I used Gemini 2 to write a Python and Selenium script for logging into an online scheduling tool that was integral to a marketing job I was doing. The test script worked, but I didn’t really take the time to understand it. After a year, I’ve fixed enough bugs with it to have a little better idea, but it’s not the same kind of understanding I used to get when writing my own (often clumsy) code by hand.
I tried my hand at using AI to help optimize a Google Ads account. And I’m not talking about Google’s own “AI-recommendations.” I mean that I described a scenario and had the AI try to give me guidance. The guidance seemed sound, if not a little off, but it was so damn confident sounding that I implemented it. But it didn’t take long to see that, no, I wasn’t sitting on the cusp of the next breakthrough for the ad account. I’d just been duped by the ultimate bullshit artist. I backed out the changes I’d made, but I wasn’t done being a slow learner. I kept using AI, because I kept hearing about how it was just so much better than all of us at reasoning. Maybe I just needed a better AI.
I subscribed to Claude and got on their expensive plan. I thought, People can’t stop raving about this. I just need to open my eyes. When it came to writing code with Claude Code, I found myself building things in minutes that might have previously taken me days. Of course, I didn’t check the code it was writing. I just kept hitting the Enter key to let it run commands. I kept refreshing browser tabs to see that, indeed, the app I was writing was being built and seemed to work. But I was also uneasy. I had no idea what was under the hood. And the thing was, all the code was write there, in a git repository, and I could look at the source code at any time with VScode. But I had become congitively lazy. My damn brain just resisted thinking. After all, if the AI could think so damn fast, and write and fix code, and all I had to do was be a “judge of taste” and think of new features, why get into the weeds? That was a red flag. In my gut, I could feel the discomfort of knowing how disconnected I’d become from what the AI was making.
I used the AI for more things. I used it to brainstorm ideas for marketing. I used it to review documents for potential legal risks before sending them to a lawyer. It seemed sort of effective. But it was also exhausting. Some of the ideas it came up with seemed great. Others were like reheated garbage, but the AI was still an enthusiastic proponent. I spent HOURS modifying markdown files, trying to guide outputs. I spent even more ideas scrolling AI subreddits, looking for things I might be doing wrong. I knew I couldn’t possibly be doing everything right, and that there was still room to “optimize” my AI stack. I heard about people using different (expensive!) cloud models for different tasks. Claude for writing code. Codex for code reviews. Gemini for reviewing entire code repositories.
Was I just not running fast enough? Was the answer to my troubles to dive deeper into the obsessed world of AI-psychosis, where if you weren’t using all the bleeding-edge AI tools, you might as well be just sipping a beer while watching people rearrange deck chairs on the Titanic? For a while, I wasn’t sure. But I felt a shift coming, at least inside me, and it wasn’t one that required a roll of toilet paper.
The Shift
My wife and I took a trip to southern France, to visit my dad and step-mom. I left my laptop at home. I spent very little time on my phone. And I spent several hours on my old road back, running a 9-speed cassette on an triple chainring LaPierre road bike that was a tank by today’s standards but still managed to deliver all over the back roads of the Tarn-et-Garonne region. I ate gluten. I ate cured sausages. I smelled flowers and walked on cobblestones. I spoke to strangers, and shared coffee with one in a café.
Slowly, ever so slowly, it felt like cobwebs were clearing from my head. After close to two years of obsessing that the only way forward as a marketer and businesses owner was to embrace AI, I realized what a mentally lazy git I’d become. And in becoming more mentally lazy (not looking things up for myself, taking AI at face value, not really diving into what kinds of outputs I was getting from AI models), I was sliding into a trap. By choosing not to think, and outsourcing my crtitical thinking, I was in fact making AI the superior thinker. By choosing not to question the endless bullshit mixed with legitimate advice mixed into AI outputs, I was becoming a slave to the machine.
So far, I’ve been talking about AI from the perspective of how it’s been impacting me. Which is inherently selfish. But as my eyes opeened further, I also started to become more aware of how AI was breaking more of us. More datacenters were being built, ruining neighborhoods, stealing water, and driving up power prices. The billionaire class were getting more entrenched and enriched, controlling the narratives around what AI would mean for humanity (and our only possible routes to salvation). AI was speedrunning human society toward Ultimate Enshittification, and each query or task I fed the machine only helped reinforce the narrative.
I’m going to speak in absolutes now. I don’t usually do that.
AI absolutes:
1) AI will never have feelings.
2) AI will never give a shit about you, your family, or their loved ones. See #1.
3) AI will never give two shits about being held accountable. See #1.
4) AI is the king supreme of generating bullshit.
5) The more we outsource our own opinions and reasoning to AI, the weaker and easier to manipulate we become.
Where to go from here if you’re a human and don’t consent to being herded like a sheep into the AI dystopia
I’m going to tell you what I’ve started doing, as a way of fighting back. Some of you may say it’s not enough. Others may say, What’s the point? And that’s okay. This is just my point of view. Here are things I think are important to reconnect with what it means to be human. To reconnect with what it means to be free.
1) Get your hands dirty. With dirt.
My wife and I planted a garden this year. It’s our first time planting a vegetable garden. But it’s also a way to become that much more independent from the machines that control our lives. With grocery prices rising, it doesn’t seem like a bad idea. We’ve got a bunch of rows of potatoes, some onions and radishes and beans, and it’s still a little too cold to put tomatoes in the ground in Sweden. It’s a start.
2) Learn to sharpen your tools.
I’ve been using Youtube in an intentional way, to learn new skills. This is not a new phenomenon. But it is a way to escape the pointless act of doomscrolling, and it’s useful. I have a water stone and an oil stone. I sharpen my chainsaw, my axe, my hatchet, and our knives. When they cut well, work gets easier.
3) Sweat.
I used to have a gym membership. After moving out into the country, I stopped using it. But instead I gather, cut, chop, and stack wood. Sometimes my wife points at lilac roots that need to go, and I swear under my breath before going out to the shed to get the mattock. Then I beat the hell out of the roots until they submit. My wife is happy. I get a workout. It’s a mutually beneficial outcome.
4) Get curious about fixing your own stuff.
I recently bought a small chainsaw, a Husqvarna 435 (and it turns out it was probably too small). I thought it seemed a bit weak in the cut, which got me to thinking that maybe I’d run it too hard. If so, it wouldn’t be anybody else’s fault but mine. So I bought a cheap compression tester at the local big box store, then realized that I didn’t trust the compression reading in the cylinder (it said 110 psi). I learned that it would have probably been better to buy a better tool from the get go. I’ve made this mistake many times already in my life.
While we’re on the topic of fixing your own stuff, I’m also grateful to my buddy CJ. He showed me how to work on my car. Which has opened up a while new world of opportunities for fixing my own shit, because my 2009 VW Rabbit is not getting younger. I’m afraid the death of it will be rust by the rocker panels, because I don’t yet know how to weld, but you never know. Either way, when you know how to do your own oil changes, flush your coolant, change your brake pads and rotors, bleed your brakes, work on your suspension, and clean out sensors and intakes, you can save yourself a whole bunch of money. It also can give you the confidence to keep driving an old car, which can potentially keep you from falling into the financial trap of having a car payment.
5) If it seems too good to be true, it probably is.
I told you I was a slow learner. I’ve bought tools that were inexpensive, thinking I was getting a good enough deal, and then the tools ended up being cheap. I’ve ended up disappointed, and sometimes having to buy new tools that are more expensive, and not saving money. I wish I could say I’ve learned this lesson for good, but I find it really hard to pass up a good bargain. However, I am in the market for a good felling and limbing axe, and I’m eyeing a Gransfors Bruk Scandinavian Forest Axe (or maybe their Small Forest Axe). You can’t really get much better than their axes. Hand forged in Sweden with a 20-year guarantee. I’m not getting paid to write that. I just really, really love their edge tools.
Also, this same principle applies to when Google representatives call you to tell you that you really should apply more of their AI recommendations to your ad account. Remember to follow the money whenever someone gives you advice.
6) Be neighborly.
I remember watching Mr. Rogers on PBS in Boothbay Harbor, Maine at my grandparents’ house at the edge of the ocean. Mr Rogers reminded us to be neighborly and friendly, and that it was okay to have a bad day once in a while. Over time, I strayed from his principles. But he was wise, and the more I remember to be a good neighbor, the better things seem to go. If you’re lucky, your neighbors can turn out to be like family you like to have around, and you can all be there for one another to help each other in times of need.
Here’s a crazy idea. I think that the billionaire-class profits from our division and fear. And I also think that when we turn to our neighbors and lend a helping hand, we become more strong and more free together.
Which reminds me, I’ve got the neighbor’s rototiller in the shed and need to take out some weeds from the tiller portion before I return it to them.
7) Remember that being human matters.
Just because Sam Altman or Dario Amodei say that the future is AI and we’ll all lose our white collar jobs and we’d better get on board their AI train doesn’t make it true. Whether they are staunch believers in their own narratives or not doesn’t matter. What matters is whether you’re willing to buy into them.
You can choose to be pro-human. You can choose what you value. You can choose to plant a garden. You can choose to help a neighbor change the tires on their car. Thinking for yourself is a choice, too.
You’re not f##ked (learned helplessness is not fate)
I like to end on a hopeful note. Maybe the list was a little bit hopeful for you. I know it was for me.
As a dad, I like to think that my boys aren’t f##ked by the world we live in, and that neither are we. Yes, there are a lot of things wrong at the moment, and it looks like things have gotten a bit out of whack with out-of-touch billionaires like Elon Musk, Jeff Bezos, and Mark Zuckerberg running amok. Not to mention the broken state of the American Republic, with its flashing check engine light entreating us all to pull over and have a good look under the hood.
Learned helplessness would be throwing our arms up in the air and saying, “There’s nothing we can change here. These forces are just too strong.” And then it’s letting the tide of so-called progress go out and sweep you away with it.
If you value humanity and being human, you can choose to be a good human. You can choose to do things that unplug yourself from the Matrix, like growing food and cutting wood. You can learn to cook a new recipe from scratch. You can get to know a neighbor and show up if they need help. You can put your phone down. You can uninstall a scrolling app. You can install a web browser that doesn’t show AI overviews. There are so many choices you can control. You can start thinking in terms of we more often, and thinking about the consequences of convenience.
(I know. No one likes thinking about consequences. It’s a bummer sometimes.)
But in thinking about consequences, maybe your choices can become more intentionally consequential. A consequence of planting a garden can be more food and spending less at the store. It can be eating food made without pesticides. It may also be a farmer’s tan at which your wife can poke playful fun.
A consequence of learning how to fix your own stuff can be saving hundreds of dollars or more on repair bills for your car or lawnmower. I can be a deeper understanding of how your machines work. It can be less money spent on buying new stuff you might not even need. It can be less stuff being trucked off to a landfill.
In my mind, we matter, and our choices matter, and we’re in this together and have the agency to make changes where changes are due. That’s not so bad, is it?
Best,
Jesse
Related on the Copy for Originals site: Should AI fully manage your Google Ads campaigns? looks at what Google’s automation and AI agents can be trusted with, and Haiku, Sonnet or Opus for online research? checks AI research quote by quote.
