ℹ️ I am rebranding and merging my newsletters. From this day, I continue writing as the Maverick Mapmaker. Because someone has to chart a course, navigating between hype and doom in AI.
Some people claim that AI coding tools will replace nearly all software developers. Others warn that the world will soon suffer a tsunami of crappy code. As usual, the hypers and doomsayers get all the attention, while the truth lies somewhere in between.
Creative Adaptation
When book publishing was democratized, many in the industry cried catastrophe. They argued that when anyone could pick up a pen or a keyboard and share their drivel with the world, professional authors would be doomed without publishers as the curators of quality. (Hint: professional authors adapted and are doing just fine.)
The same happened with professional photography, videography, and web design. Whenever new technologies democratized creativity and opened the floodgates, professionals had to adapt: seek higher-value work, diversify into adjacent domains, and fix the mess that unpaid amateurs created.
Whenever new technologies democratized creativity and opened the floodgates, professionals had to adapt.
The Photography Case
Take photography, for example.
The transition to digital cameras was evolutionary for professionals, but the rise of smartphone photography represented a revolutionary disruption. It disproportionately affected the lower end of the market, where technical requirements were minimal. And let’s be honest, I doubt many passport photographers are mourning the loss of that kind of work.
Professional photographers shifted their focus to premium services requiring specialized expertise and equipment. Photography is more than just ‘point and shoot.’ Professionals spend years mastering the technical aspects of shutter speed, lenses, lighting, composition, and more. This expertise remains valuable, particularly in high-end markets.
Meanwhile, some photographers expanded into adjacent areas, leveraging innovations like drone photography, underwater photography, or microscope photography, which made previously inaccessible shots possible and expanded creative opportunities.
And then there are those who make a little money on the side fixing the terrible photos Uncle Jeremy took at last week’s wedding anniversary for Aunt Jane. Those photos would never have existed if Uncle Jeremy hadn’t convinced Aunt Jane that his iPhone 10 could still do a marvelous job. When you clean up someone else’s crap, you might as well charge a modest fee.
(Now, replace photos with apps and you see what will to happen.)
The Creative Tsunami Effect
Thanks to the democratization of content and code creation, the amount of data in the world is rising exponentially. Most of the texts, photos, and videos that ever existed were produced in the last five years.
Few people realize that opening the creative floodgates has multiple compounding effects:
- Low-value work shifts to amateurs because they will do it for free. (For example, in 2011, my spouse and I asked a few friends to take our wedding photos. The pics turned out great, and we spent the money we saved on better food for all our guests.)
- Amateurs swarm over niche markets that professionals never addressed because they were too expensive (the Long Tail Effect). You won’t find many professional photographers covering your local fitness club’s 100-meter race for toddlers.
- Automation creates more complex tasks as human work shifts toward higher-skill activities, not lower (the Automation Paradox). For example, autopilot hasn’t replaced airline pilots; it shifted their role to complex problem-solving when automation fails.
- New tech enables high-end markets for professionals to offer services that were previously too expensive for everyone. (A rising tide lifts all boats, whether the captains are amateurs or pros.)
- Lower costs increase demand (Jevon’s Paradox), meaning both amateurs and professionals benefit from the same tsunami. As the market grows, even a shrinking percentage for professionals can still mean more work in absolute numbers.
- Professionals fix the work of amateurs because ninety percent of everything is crap (Sturgeon’s Law), which means professionals will have plenty of work cleaning up the mess after amateur mistakes.
- Scarcity increases value. The more something becomes a commodity, the more customers pay for artisanal, high-quality work (the Scarcity Principle).
The Next Flood: Apps
With vibe coding tools like GitHub Copilot, Claude Code, and Cursor, the floodgates of code are now opening. Anyone with a systems-thinking mindset can predict what happens next in software development.
First, more apps will be created than ever before. The amount of code written in the next decade will likely dwarf everything written so far. And that means:
- Low-value coding will be free. I doubt many engineers will look back nostalgically on that online reservation form they built for the neighborhood’s karaoke bar.
- Most new code will target niche markets. Expect custom apps for local football clubs, next-door hair salons, and Grandma’s 100th birthday party roulette.
- The human role shifts from simple to complex. As I wrote in my book, human work is shifting from tame (solvable) problems to wicked (unsolvable) ones—meaning more time spent architecting, designing, and managing stakeholders rather than writing boilerplate code.
- AI unlocks new opportunities. Experienced engineers can build complex systems that were previously uneconomical to create.
- Lower development costs increase demand. More code means more software dependency, meaning more work for both amateurs and professionals alike.
- Most AI-generated code will be bad. Ninety percent of everything is still crap, meaning professional engineers will be in high demand to fix everyone else’s problems.
- Even if AGI writes perfect code, top-tier software engineers will still thrive. The artisanal coders at the high end of the market will command even greater value.
Point six is particularly relevant because, unlike bad books or photos, bad software can’t just be ignored—it has to be maintained. When a school teacher builds AI-generated learning games that become wildly successful, they’ll eventually hit a wall where the AI lacks product vision, planning, and maintenance capabilities.
Additionally, AI is great at writing code but is still terrible at debugging it. AI-generated code can be subtly broken in ways that are difficult to detect, producing convincing-looking garbage. This is a major risk for mission-critical systems requiring security and reliability. Unlike bad photos, bad software can’t be discarded once people have come to depend on it.
More apps will be created than ever before. The amount of code written in the next decade will likely dwarf everything written so far.
Conclusion
Anthropic CEO Dario Amodei recently claimed that by March 2026, nearly all software will be written by AI. That might be true, but most of that will be code that no human engineer would have written.
With 1,000 times more apps being coded in the next five years, we will likely have more professional developers, not fewer.
And as AGI approaches, some organizations may proudly start displaying the names of their human engineers: “This software was lovingly crafted by our human developers [insert your names here], supported by their AI copilots.”
I did the same with my book.
The AI Debate is Stuck. Let’s Map a Smarter Route. Drowning in AI hype? Tired of AI doom? With my work, I aim to chart a third path—skeptical, but forward-thinking, steering between evangelists and apocalyptics. 👉 Subscribe now to get my take on the future of work.
