AI was Killing My Career. What To Do?


I’ve spent my adult life making money at writing. Not always a lot, but usually enough. That’s been especially true since around 2008, when I left a longtime career in journalism (Seattle correspondent for The Economist, contributor to Outside, Travel + Leisure, Sports Illustrated, others) and went corporate. Since then, I’ve written blogs, articles, white papers and more for Microsoft, Amazon, Alaska Airlines, KPMG, and other prominent companies.

Early in 2023, I began using AI (artificial intelligence, for anyone living under a rock) as a writing assistant. I found it incredibly powerful – so long as I put careful boundaries around what I wanted it to do. I could take copy from a PowerPoint deck, a raw transcript from interviews with subject-matter experts, maybe an example of a previous piece, feed it all into ChatGPT, give it advice on tone, audience, and format, and let it rip.

At the very least, I’d get a first draft I could build on. Sometimes even a draft that didn’t require a ton of work at all. Memorably, I once wrote a piece for a client titled, “Will AI Be Your Next Internal Auditor?” Or rather, I had AI write it based on a conversation with people from the client company, plus my own suggestions about covering the “black box” problem (the idea that no one really knows what’s going on inside the AI “box”). With a few minor tweaks, I sent it to the client, who liked it…a lot. They then asked me for a Twitter post, a LinkedIn post, email copy, and web copy, all to drive traffic to the site.

I had AI do that, too. And the only giveaway that it wasn’t me is that AI was MUCH more thorough – for instance, adding twice the hashtags I typically would.

Overall, AI made me more productive and very often resulted in clients who (unknowingly) preferred the AI-assisted pieces to those written entirely by me. Plus, it was fun to use. I figured it would add several years to my working life (I’m old enough to retire but have chosen not to).

Wait a minute. Who is taking my work?

But as 2023 went along I noticed I had fewer requests coming in from new or existing clients. That eventually led to a stretch from November to February of this year when I landed NO new work. That was unprecedented. I had been consistently busy – and often swamped – for the past decade. Even COVID resulted in just a small hiccup before clients needed work that showed their reaction to the pandemic.

As all of this was going on, I happened to be finishing work on a long-time passion project: the renovation of a 1979 Airstream TradeWind, a 25-foot trailer. I also had stashed away a 1977 Overlander – a 27-footer – with the intent of starting a renovation of that, along with setting up a YouTube channel of videos on the process. For fun, I acquired a business license, a domain (, and some branded merch. Hats, totes, T-shirts, that sort of thing. I also did a few small jobs, strictly through local word of mouth.

So, earlier this year, as it finally dawned on me that I was becoming a victim of the very AI that I was using, I decided I had to make a serious effort to monetize what I had acquired both in knowledge of Airstream renovation (I literally, albeit ironically, could write a book) and in tools (metalworking tools, welding machine, and on and on).

I bought a small ad in Vintage Camper Trailers magazine, built out my website, launched a Facebook page (I loathe Facebook but really didn’t have a choice), started advertising on Facebook and Google, and began work on YouTube and Instagram videos.

First, the merch (Image: Doug Gantenbein).

Do I have a new, profitable vocation? That’s not clear so far. It looks like I may have a client who needs about $30,000 of work on a ’68 GlobeTrotter, and I’ve been getting leads from Facebook. I also know that the sum of people doing similar work in Oregon, Washington, and Idaho is exactly this: three, including myself. Given the longtime popularity of Airstreams (quick, name one other RV brand) and the fact that even here in Port Townsend I know of about a dozen old Airstreams that could use work, I have to believe the market is out there.

More to the point, the market is as AI-proof as anything I can think of. True, I’m now using AI a little to do things like script out a YouTube series. But in terms of doing the work? Me and my two hands – they will be doing the actual work.

Can we AI-proof ourselves?

I think my change of course will be a trend writ large. Article after article I’ve come across recently lists the jobs AI is going to take: computer programming, financial services, graphic design, content, and on and on. I firmly believe AI will do a lot of good in the world – in fields such as healthcare, for instance. But it also will kill a lot of careers.

I know – lots of articles on the web have suggestions on how to AI-proof yourself. Here are some of the tips: “Avoid predictability.” “Embrace your humanity.” “Develop emotional intelligence.”

To which I say: Really? I mean, really??? Please.

What to do? One word: trades. For instance, I’m soon going back to school at Olympic College in Bremerton. I’ll be earning a certificate in aluminum welding, which requires a particular technique but is applicable to welding any metal. Now, for the most part, this is to support my Airstream work – frame repair, fabricating wheel wells, that sort of thing. But in theory, my new skills also will qualify me to walk into a nearby boat shop that builds specialized aluminum boats and ask to be an apprentice. Which I very well might do.

As a report from the World Economic Forum points out, “AI is unlikely to be able to replace jobs requiring human skills such as judgement, creativity, physical dexterity and emotional intelligence.” The report goes on to list the jobs it deems most AI-resistant. Unsurprisingly, of the top six, only one is not a construction or heavy-equipment job.

Welding, plumbing, electrical work, carpentry – these are the sorts of things that AI will never replace. Anything that requires bespoke work and the judgment to understand how to approach tasks that are infinite in their variety. Like renovating vintage Airstreams.

Should everyone become a welder? Probably not, although one could pretty easily find work sufficient to earn six figures, or close to it. Welding is welding – whether one is working on an Airstream, a skyscraper, or a piece of jewelry, the technique is the same. So it’s something of a universally needed skill. There is debate over just how viable welding is as a career option, but much of that debate is pre-AI.

Prospects are good for nearly any trade. Every plumber, electrician, and carpenter I know has all the work they wish, often in the neighborhood of $80 per hour plus the cost of materials. That’s potentially north of $150,000 a year – twice the U.S. median.

Not a solution. But a decent option

Obviously, if everyone becomes a tradesperson, the demand will collapse. I’m not suggesting that (or maybe I am…). But for people like me who have made a living with our heads more than our hands, it’s something to think about.

Moreover, endeavors such as Airstream renovation are, to my way of thinking, the perfect combination of art and science. It requires me to think about design, structural integrity, electrical loads, solar power, HVAC, and a million other intricacies. It’s the handwork process that I find particularly appealing – creating something tangible and physical, not digital. Same could be said for many trades that are often downplayed as such.

Whether or not we become a nation of plumbers and welders, I certainly think we ought to ensure that people know how to work with their hands. When I was in the 7th and 8th grades at Capitol Hill Elementary School in Portland, Oregon (1967-68), boys were required to take shop class. I learned basics of wood shaping, soldering, and other skills, ones I occasionally draw upon even now. True, the girls were shunted to “home economics,” which was wrong, but even that had practical applications, which boys would benefit from as well.

Put every kid through both. Why not? Swinging a hammer and knowing how to whip up a French omelet are skills anyone can use. Clearly, over the past two decades we have put far too much emphasis on four-year colleges and universities. And as we are seeing, a four-degree is hardly a ticket to guaranteed wealth. I mean, let’s say you graduated with a degree in computer science two years ago. Now what? More than 191,000 workers at U.S.-based tech companies were laid off in mass job cuts last year, and the layoffs continue.

We used to be a nation of people who built things and fixed things. Indeed, that was a competitive advantage during World War II, when our military was stuffed with 20-year-old farm boys who could fix a 14-cylinder Wright Cyclone aircraft engine with duct tape and a wrench. That was huge compared with our adversaries, Japan and Germany, whose people were almost entirely lacking in these skills (Germany put much more emphasis on craft rather than mass assembly and repair).

So, pick up a welding iron. Or a radial saw. Or a plumbing wrench. Make yourself AI-proof.

Douglas Gantenbein
Douglas Gantenbein
“Douglas Gantenbein has worked as a journalist, writing coach, and marketing writer. He lives in Port Townsend with his wife Jane – that is, when he is not hiking in the Grand Canyon.”


  1. Love that you found a way to adapt to the changing labor market. I have teens contemplating college right now and I ask myself, how in earth do we plan a life at this juncture?

    • I have no idea. What do you tell them? Coding (nope)? Financial services (nope)? Graphic design (nope)? It’s impossible to know. The only jobs that will be AI-proof, in my opinion, will involve things where every project is bespoke — and often has a manual component.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Comments Policy

Please be respectful. No personal attacks. Your comment should add something to the topic discussion or it will not be published. All comments are reviewed before being published. Comments are the opinions of their contributors and not those of Post alley or its editors.