AI Isn't Coming for My Job. I'm Coming for Its Job.

Follow me back in time for a minute.

Where were you when you first felt it? When you tried an LLM for the first time and thought:

What does this mean for us?

The first week ChatGPT came out, I hadn't gotten that far. I asked it to write a funny script. I had it tell me jokes. I played 20 Questions. It was all fun and games. Literally.

I remember being astonished by it, but I don't think my brain could really conceive of what had just happened. I was talking to a computer, and it was talking back. Not in the way computers had talked back before, with their error messages and carefully programmed responses. This thing seemed to understand what I was saying.

So naturally, I asked it to tell me another joke.

That was almost four years ago. And I think it's fair to say things have gotten a little weird since then.

These days, artificial intelligence is a political argument, a business strategy, an existential threat, a career opportunity, and an apparently inexhaustible source of videos of Tom Brady krumping.

Depending on whom you ask, it's going to cure cancer, destroy democracy, eliminate poverty, consume all our drinking water, or make every human being on Earth unemployable by next Thursday.

We're expected to have opinions about all of this. Strong ones, preferably. We're supposed to know which tools to use, which companies to trust, what skills to learn, and how to prepare for a future that changes every time we get comfortable with the last prediction.

And meanwhile, we still have to go to work. Pay our bills. Buy groceries. Be people.

Jesus Christ! We’re tired.

Somewhere in all the arguments and memes and product announcements, we've lost something.

This is an incredibly profound and angsty and scary and beautiful thing that we are all experiencing together.

For all the conversations we've had about AI as a harbinger of God knows what, we never really stopped and recognized it for its emotional weight. We never processed our excitement, our fear, or our grief. We bulldozed past it, and we tried to rationalize it.

We're excited and afraid. We're curious and resentful. We want to understand what's happening, but we're tired of hearing about it. We're afraid of being left behind, and sometimes we'd really like everyone to shut up about the future for five minutes so we can enjoy the present.

And of course, we aren't all experiencing this equally. Some people have already lost jobs. Artists have watched their work become training material without their permission. Communities are being asked to bear the environmental costs of infrastructure they had little say in building. Those things are real. They deserve more than sympathetic words. They deserve action, accountability, and regulation.

But there's room, alongside all of that, to admit that we're having a strange collective emotional experience. We don't all agree on what should happen next. We don't even have to agree on what we're feeling now. We're figuring it out together.

I certainly haven't had a particularly graceful time of it.

I'm a writer. I studied writing. I've spent my career getting better at writing. So when a machine showed up that could produce a passable essay in fifteen seconds, I had some feelings about it.

And those feelings weren't entirely irrational. AI has genuinely made it harder to find certain kinds of writing work. It's changed what clients expect, what they're willing to pay for, and what they think they're buying.

I liked how much faster I could research something for work, but hated the AI-generated songs that started showing up in my feed. I loved that I could get through a writing assignment more quickly, but I did so with something like the feeling I imagine GameStop employees have selling digital PS5s: Sure, I'm making a buck today. But I have a question or two about tomorrow.

At times, I'd see some new image or video generation demo and scroll right past it. I actually wanted to avoid seeing what had become possible. Every new breakthrough came with another little reminder that the world I knew was changing, and I couldn't do much about it. I wanted to find somewhere cozy and familiar and stay there for a while longer, where everything was beautiful and nothing hurt.

Of course, you can't really do that. One day you walk outside and the familiar just isn’t there anymore.

There's a sadness in that. Perusing the aisles of Blockbuster. Tuning in live for the Friends finale. Staying up with your friends for midnight releases. Experiencing something at the same time and in the same way as everyone around us.

I miss those things. I don't think we have to pretend every inconvenience of the past was secretly wonderful, or that every change has made our lives better. Sometimes the things we lose really were beautiful. Sometimes they brought us together in ways we didn't appreciate until they were gone. But they remain beautiful because of what they meant to us then. We don't have to preserve the world exactly as it was for those experiences to have mattered.

I think I knew that intellectually. Emotionally, I was having a harder time.

For a while, my relationship with AI settled into something practical. I used it to research, brainstorm, organize information, and help with writing. Sometimes it was impressive. Sometimes it was infuriating. Sometimes it gave me a really wonderful answer to a question I hadn't asked.

I'd read people talking about the importance of prompting, and how AI was only as good as what you put into it. Garbage in, garbage out. AI is what you make of it.

And I'd think, well, what the fuck does that mean? I'd tried it. I'd gotten pretty good at using it. It helped with some things. I could see why people liked it. But I felt like I'd found its upper bound, and frankly, it wasn't that great.

Then I started approaching it differently. I was trying to tell a story, and I had a very specific idea of how I wanted it to feel. I wanted it to read like an episode of Dateline.

I'd watched plenty of Dateline. I knew what it felt like when a detail introduced ten minutes earlier suddenly turned the whole story on its head. I knew how a revelation could change everything you thought you understood. I understood the experience from the couch. I had no idea how to engineer it on the page.

At first, I did the obvious thing. I asked AI to write it. And the results were fine. That's the frustrating thing about a lot of AI writing. It can produce something with all the recognizable features of what you asked for. The suspenseful opening. The transitions. The big reveal. Everything is where you'd expect it to be, and somehow none of it is working.

But I didn't want fine. I knew what I wanted, and this wasn't it. So I started taking the story apart.

What did the reader know at this moment? What were they supposed to be wondering? What happened if I moved a detail three paragraphs earlier? Would the reveal still land? Why did one arrangement make me want to keep reading while another felt like homework?

I was engineering the story from above, figuring out how the components worked together instead of trying to write it from beginning to end. I could use AI to explore the structure, identify what wasn't working, and try again. I still had to judge the results. But now I had a way to work on the parts of the craft I didn't know yet.

And suddenly, something clicked. I didn't have to know how to do every part of something before I could begin making it. Now I couldn't help but wonder what else I could do that with.

Until a couple of months ago, I'd never coded anything. Software development wasn't a skill I possessed, or even one I'd seriously considered acquiring. Now I'm building software.

To be clear, this involves a great deal of commitment, ingenuity and trial and error. I am constantly discovering new things I don't know. AI makes mistakes, sometimes spectacularly, and I'm learning that getting something to work once is very different from getting it to work reliably.

But I'm doing it! I'm making things I wouldn't even have considered attempting a few months ago.

And the funny thing is, once you start doing that, you begin to realize how many ideas you've dismissed over the years without even giving them a second thought. Things you wanted to make, questions you wanted to answer, problems you wished somebody else would solve. You didn't know how to do them, so you moved on.

I'd been asking the wrong question all along.

I was asking: How can I use AI to make the things I already do easier?

I should have been asking: How can I use AI to do things I couldn't do before?

I used to watch Project Runway and wonder why all the designers had to be able to make their own clothes.

Sewing is an extraordinary skill, obviously, and the people who do it well bring their own creative intelligence to the work. But designing a beautiful garment and physically constructing it are two different abilities.

What if the best fashion designer in the world never got the opportunity to find out because they couldn't sew?

Or imagine someone with a genuinely original idea for a movie.

They're sitting in a diner in Los Angeles at three in the morning, scribbling the plot on a napkin. They can see the movie in their head. They know the characters, the story, the moments they want people to remember.

But that idea doesn't come with a lighting plan. It doesn't tell them how many extras need to be in the background or what the sound design should be. It doesn't account for the thousand creative and technical decisions between imagining a movie and actually making one.

Those decisions matter. Great cinematographers, actors, editors, and sound designers can make a film into something its original creator never could have imagined alone.

But what happens when the person in that diner can start experimenting with those decisions for themselves?

When they can actually make a scene, see what works, change it, and keep going?

How many wonderful ideas have never seen the light of day because the people who had them couldn't afford to find out if they were any good?

For most of the history of personal computing, if you wanted to make a computer do something genuinely new, you had to learn to speak its language. You needed specialized training, or the money to hire somebody who had it.

Now you can start by explaining what you want in your own words.

You can sketch it. Draw a diagram. Ramble incoherently about it for twenty minutes. You can see a picture of a fire hydrant that inspires you for reasons you can't quite articulate and use that as the starting point for something.

The computer might misunderstand you. You might have to explain yourself six different ways. You still need to make decisions, learn things, and figure out whether the result actually works.

But you can begin.

And once you realize that, it's difficult to stop thinking of things you might want to try.

For a long time, technological progress felt like something that happened to me.

A company would announce a new product. A platform would change its algorithm. Some new capability would appear, and I'd have to figure out what it meant for my career, my hobbies, or the world I lived in. And like a lot of people, I had complaints. I still do.

And I want to be very clear about something, because I don't want anyone mistaking my point here for endorsement of the companies behind these products.

OpenAI and Anthropic do not deserve your blind loyalty.

These are extraordinarily powerful companies developing technology that affects all of us, including people who have never used their products and have no interest in doing so. They have a social responsibility to you and to me. They deserve scrutiny, criticism, and accountability.

And our government has a responsibility to protect our interests regardless of how many lobbying dollars these companies spend.

I want strong regulations. I want meaningful protections for workers and artists. I want communities to have a say in decisions that affect their lives and environments. I want the people building this technology to be answerable to the public, and I want our elected officials to remember whom they were elected to represent.

I don't think we should surrender any of those demands. But I want to encourage you to set that argument aside for just a moment. You can pick it right back up. I'll be right there with you.

For now, I want you to approach this from the other side. How can this empower you?

If something bothers me, I can investigate it. I can examine public information, look for patterns, and try to find evidence that might help somebody do something about it. If I think a particular kind of software should exist, I can try making it.

I might fail. I might discover the problem is harder than I thought, or that my original idea was terrible. Some problems are far too big for an individual to solve, no matter what tools they have. There is no replacement for the power of collective social action.

But I don't necessarily have to leave every problem, every idea, every possibility to somebody else anymore. And I don't think you should have to either.

Here’s the immutable fact: Change is coming. You can fight it, you can salivate over it, you can bury your head in the sand. But it’s coming. And you deserve to have a say and a hand in what it looks like when it gets here.

That doesn't mean everyone has to become a software developer or an entrepreneur. God knows we have enough people on LinkedIn telling us to turn our hobbies into revenue streams.

You don't have to change the world! Maybe there's a little inconvenience in your daily life you've always wished somebody would fix. Maybe you want to make a short film, or a game, or an illustrated story for your kids. Maybe there's something in your community you've been curious about for years, and you'd like to understand it better. Maybe you just want to make something completely ridiculous because the thought of it makes you laugh. That's enough, too.

One of the most radical things about what I've been experiencing lately is that I've started having fun again. That wild, unbridled toddler kind of fun where you get an idea, begin messing around with it, and suddenly discover that several hours have passed. I hadn't realized how much I'd missed that feeling.

So here's what I'd suggest. Forget, for a moment, about optimizing your workflow. Forget about becoming an AI expert, keeping up with every new release, or finding some way to monetize this astonishing new capability.

Think about something you've wanted to do but never really considered possible. It doesn't have to be profound. Start with something small. Something you'll enjoy working on even if it never becomes anything more.

Open up an AI tool and tell it what you're imagining. Tell it what you know, what you don't know, and what you'd like the finished thing to look or feel like. Then ask it to help you figure out how to make the smallest possible version.

Don't expect it to get everything right. Question its answers. Test what it makes. Change your mind. Ask why something isn't working. Try, fail, learn as you go.

I can't promise you'll make exactly what you imagined. There are still plenty of things these tools can't do. But you might discover that something you thought was out of reach is a little closer than you realized.

I spent a long time worrying about what AI was going to do to my future. I'm still worried about parts of it. I think we all have good reasons to be. But lately, I've been thinking a lot more about what I'd like to do with it.

The magic isn't really magic. But it can be.