Why AI Will Kill Your Job So You Can Make a Better One

And will someone please start the YouTube of user+AI-generated software?

I believe that the existential worries about AI putting us all out of work are misguided, but the reasons are different from what almost everyone seems to be talking about.

It starts with this: People want to work. Sure, some don’t, but most do, and that applies even if the gung-ho AI fanboys are right when they say AI will create some kind of utopian abundance and spin off so much wealth that we’ll all get a sizable monthly AI dividend.

Most people would still want to do something constructive, and a lot of people would still want to make money, even if they get a dividend. Just look at scions of super-rich people. They don’t have to do anything but play, yet most of them choose to do something at least somewhat purposeful. (David Ellison certainly doesn’t NEED to try to run the world’s biggest media company…)

So if AI starts booting huge numbers of people out of the kinds of jobs they’d been doing, that means AI would be creating a gigantic, important, urgent problem for huge numbers of people. It’s both a personal and societal problem: How can people do meaningful, remunerative work in the post-AI era?

In a long career of writing about the tech industry and a decade of working with companies on category creation and design, one thing that’s become super clear to me is that gigantic, important problems are magnets for entrepreneurs and investors. New categories of products and services are born when someone sets out to solve a problem.

The bigger the problem, the more valuable the category and the companies in it. Both the challenge and the potential wealth creation are the major attractions.

Now, my business partner, Mike Damphousse, and I have a new book coming out based on our work, titled The Category Creation Formula. This is relevant to the AI and jobs problem. The “formula” is simply: context + missing + innovation = a new category. As a way to help a company’s leaders see, frame and build a new category, the formula as a discussion driver works every time.

Briefly, context is what’s going on in the world, which includes new technologies, shifts in geopolitics, economic conditions, changes in societal mores, etc. The missing is an unsolved problem in that new context – maybe it’s a new problem, or maybe it’s an aching old problem that now can be solved because of the context shift. Finally, the innovation is the solution to the problem. Create that innovation, and you’re creating a new category of product or service that the world needs and the market will vacuum up.

The bigger the context shift, the bigger the new problems that need solving. And the bigger the problem, the more powerful the magnet compelling entrepreneurs and investors to it. If AI puts hundreds of millions or billions of people out of traditional work – that’s about as big a problem as you’re gonna get.

The usual optimistic counter-argument to “AI will kill all the jobs” is that new technologies always make some kinds of work obsolete, but create entirely new kinds of work. And, yes, this has been true. My grandfather was an ice deliveryman in the era of ice boxes in homes. Electric-powered refrigerators killed that job in the 1930s. But then electricity opened up a new career for him, as an electrician. If my grandfather were around today, he’d be astounded that anyone could have a job doing something called “search engine optimization,” entirely created by 2000s-era technology. Career paths have always constantly morphed, and generally technology has made work better and more abundant for society as a whole.

Besides, if new tech always killed off old jobs but didn’t create new ones, by now we’d all be living the life of a house cat.

The Elon Musks and Sam Altmans say that this time it’s different. AI is going to become so powerful so fast, it will be like a tsunami wiping out much of what people now do for work. Whatever new jobs it creates will be done by AI or robots! We’ll be left on the sidelines.

And, yeah, it can be hard right now to foresee all the many wonderful new jobs for humans that AI will create. So, what could possibly save us?

But…people want to work, and having billions of jobs killed by AI is a problem many ambitious people will want to solve – for society or for themselves. I believe we’re seeing the beginnings of how that will play out.

Early this year, Anish Acharya, a partner at Andreessen Horowitz, published a newsletter titled “Software’s YouTube Moment is Happening Now.” He pointed out that before YouTube (and other enablers that arrived around the same time, like cameras on phones and video editing software), video media was the province of professionals. Everyday people didn’t have the equipment or specialized knowledge to produce a show, movie or music video, and even if they did they didn’t have a way to put their work in front of millions of people. A relatively small number of people had careers in television or movies.

Technology changed that. Fast-forward to recent years, and YouTube reaches an audience of billions for free, and then there’s also Instagram and TikTok and other platforms. Great phone cameras are ubiquitous; editing tools are free and simple. What used to be costly, hard and took professional training today is open to almost anyone with an idea. There are now an estimated 27 million paid content creators in the U.S. alone. That’s 27 million jobs that didn’t exist before – jobs that no one in, say, 1990 could’ve imagined.

Acharya argued that code-writing AI, like Claude Code, is ushering in a YouTube moment for software creation. “If you’ve never programmed, you can type into a text box and see what happens,” he wrote. “Today, anyone can ship an app.”

In January Steven Strauss, a professor at Princeton and Harvard, made a similar point based on his direct observations: “Last summer, student teams of mostly non-coders – marketing professionals, operations managers – built minimum viable products in days. Not mockups. Working software. Following a vibe coding workshop, students in my Harvard Summer School course accomplished things that would have been unthinkable two years earlier.”

The ramification of what both Strauss and Acharya are saying is enormous. Over the past couple of decades, technology has turned content creation into millions of jobs. Software creation is even more powerful. Software does things other people need to have done. It can be a business. AI is making it easy and inexpensive to build software products. AI will turn business creation into something that millions of people – people who never had a path to building a tech business – can do.

“If you run an incubator, teach entrepreneurship or management, or make early-stage investment decisions, the research consensus is now clear enough to act on: The economics of launching a venture have fundamentally changed,” Strauss wrote.

“Software is becoming a viral medium, and soon no one will have any excuses not to make it,” Acharya concluded. “There’s a lot of angst about ‘the kids’ these days. I have no angst, but maybe a little envy. AI has massively democratized leverage and productivity for creative people. There has never been a better time to be a young person with great ideas.”

However, there’s still a huge problem to solve. To stick with the YouTubification analogy, Claude Code and its ilk are more like the phone cameras and easy editing tools of a decade or two back. They made it easy for anyone to create video content. Coding AI makes it easy for anyone to create a software business.

What’s missing for software and software businesses is a software version of YouTube. YouTube gave creators a way to distribute what they created to the masses, and eventually gave them a way to make money on what they created. Other distribution platforms followed – Instagram, TikTok, and similar platforms around the world.

Today, an open, free, easy-to-use online mall of individually-built software businesses would probably be a great unlock for new-era jobs for millions. That seems like one huge category-creation opportunity that I’m betting many founders will chase.

What will that innovation be like? Too hard to know. Soon after YouTube was founded in 2005, I was at a tech conference where two of YouTube’s founders, Steve Chen and Chad Hurley, showed off a service that at the time was meant to let people upload home videos so friends and family could see them. That by itself seemed miraculous. No one in that room foresaw it would become a $50 billion business and one of the world’s most powerful media outlets. Anyone who thinks they know how a user-generated software business platform will evolve over the coming years is just guessing.

People want to work. AI might make your job obsolete, or make it harder to get hired because AI can help one person do the job of ten. A company that used to employ 300 might only need 30. But a lot of people who can’t find a way to do what they used to do will have a path and motivation to start and run their own one-person or five-person company.

Multiply that over and over and we could end up with a different kind of economy – one where big companies and big company jobs get more rare while there’s an explosion of very small and profitable companies. (Which, by the way, could be based anywhere.) The nature of those AI-enabled companies and jobs will be as foreign to us as a search engine optimizer or Instagram influencer would’ve been to my ice-delivering grandfather.

In 2018, Hemant Taneja and I published a book titled Unscaled, based on our theory that twenty-first century technology – AI in particular – is going to take apart the scaled-up, mass-market, mass-production industries that were built in the twentieth century. Giant companies that sought to make the same product for the most people would give way to millions of small entities that profitably make exactly what narrow niche markets desire. I think the Youtubification of software is an accelerant for that trend.

Is AI going to blow up a lot of careers? No doubt. Will it create new kinds of great jobs for humans – jobs humans do because AI can’t? Very likely, just as electricity killed off ice delivery but made way for an electrician who could wire up a house for a refrigerator.

But the more exciting probability is that AI will give us humans a kind of agency we never could have envisioned. AI will give humans a way to dream up and create their own jobs, their own new kinds of work, and their own now-unimaginable careers.

People want to work. I believe they will.

—

This story ran in the Binghamton Press on December 16, 1937. The man on the left is my grandfather. Also, he liked highballs.