Why Every Desk Job Is Becoming An AI Apprenticeship
I don’t write code anymore. I’m a software developer, and writing code used to be a big part of my job. Now AI does it for me. What used to take hours takes minutes. And this isn’t just happening to developers, it’s coming for every desk job.
The Cost Of Writing Code Approaches Zero
In software, we write designs before we implement. The same is true with AI agents. They submit plans for humans to read before executing. It takes me just as long to read the plan as it takes an AI agent to complete the coding task! What used to be the most expensive part of my job, taking hours to days, is now made trivial with AI. This means that software is becoming cheap! But it has also thrown many developers into crisis, as the years of post-secondary education are in question.
We All Shall Fall
Developers are rightly questioning the value of their education in a world where AI can accelerate computer-related tasks significantly. AI isn’t just good at writing code. It excels in all kinds of workflows. I recently consulted with a YouTube expert on how to automate his workflow. In his case, he wanted to automate finding performance outliers in niches for his clients, to identify trends and likely winners. This is a no-code task, proving that AI’s vast knowledge and skills can be unlocked in computer-related tasks across many disciplines.
Questioning Post-Secondary Education
When AI has the knowledge and skills, what matters most is having a curious mind, and the willingness to learn. What then is the point of an extended post-secondary education? Surely the time kids spend in school should be enough to foster a curious mind and the drive to succeed. In today’s world, with AI’s help, the right frame of mind should be enough for many desk workers.
For many years, apprenticeship has been the way of the trades: plumbing, carpentry, welding, electrical work. The education period is much shorter, followed by years of on-the-job training. The same will be true of knowledge work going forward. There is very little point in pursuing years of post-secondary education, when knowledge workers, bootstrapped with a short training period, have the skills to be immediately effective on the job with access to an AI agent.
The Future, Now
AI is growing at a rate that outpaces humans’ ability to adapt. Anecdotally, recent developments in AI have resulted in at least a 10x multiplier for me. But businesses don’t respond immediately by growing 10x in a month. Unfortunately, to realize the gains, the easiest solution is layoffs. If a developer can do the work of 10 other developers, then reducing the headcount is a quick way to realize the gains. The impact of this is currently being felt in the market for developers, as many are unemployed and have experienced significant disruptions in their careers.
What this means is that it takes time to adapt to this new reality. AI will be the new reality for the next generation, and likely they will grow up in a world where AI tutors will be a major component of desk work. But make no mistake, the transformation is already underway. Many developers, myself included, see no point in writing code manually any longer. For us, what matters is asking the right questions to unlock AI as a productivity multiplier.
Conclusion
Apprenticeships will be the new reality for all workers, not just the trades. Software developers are undergoing a significant disruption to their careers due to AI, a phenomenon that will soon spread to all desk workers, though the full realization of this phenomenon is likely a generation out. We will have to rethink the value of a years-long post-secondary education in preparing the next generation for the workforce.
It’s my opinion that this reality is better embraced than resisted. I really enjoy using AI for my personal projects and workflows, but I too have suffered from displacement in the job market. Like all technological innovations of the past, workers do get displaced, but humans adapt and new opportunities open up that were unforeseen.
What do you think? What does the future of knowledge work look like? Comment below ⬇