← All Posts
The word on the street these days is that AI is taking over your job, doing it faster, cheaper, and better than you ever could. And you’ll be out of a job soon!
Let’s break it down and see how you can survive and thrive.
Imagine you’re a knowledge worker like a programmer. You take a story from the backlog and implement it. What do you do next? You take the next one. You’re only done when your backlog is empty. An empty backlog is a mythical concept that’s never been achieved (at least so far).
When do you lose your job?
AI is helping you complete your tasks 10 times faster – you are able to do in a month, what used to take a year.
However, your backlog could be empty if the people adding to it can’t keep up. These are usually founders and managers who are making bets based on the company strategy.
Can AI help them do this faster? Yes. Could they build it themselves instead of asking you? This is where the trouble starts.
If they get it done good enough, much faster and there really isn’t much you add, you are the first to go. New graduates might struggle here. (There is another perspective that says a Junior programmer can now ship as well as a senior programmer because of AI - and they take a much smaller salary. But I don’t buy that train of thought because it’s short term)
If you can get AI to implement faster and better than the person assigning tasks to you, you might have a bought some time. This is a person also great at reviewing code and speaking the programming language in abstract terms. For instance, you could suggest AI to “use the strategy pattern to implement the model gateway” or know the right tool for the job “using a private cache with SWR” . This skill allows you to guide AI to achieve the desired outcome and ensure accuracy faster. However, as LLM providers improve their models, they will likely incorporate this capability as well.
If you have high agency and business acumen, you’ll have longer. Instead of working based on assigned tasks, you can seek out work and not be at the mercy of the person building your backlog. For instance, if you know your customers need higher accuracy in data classification, you can work on better datasets, algorithms, or strategies without being asked. Proactively setup monitoring to report errors and performance issues and fix them.
If you’re someone who can handle multiple steps in a workflow, such as product strategy, UI design, and programming, you can become incredibly valuable. AI can enhance your value by a factor of ten:
By protecting your time, you can work in a flow, like water finding the path of least resistance and through the cracks instead of trying to hammer through a mountain. If you discover that the original approach doesn’t handle certain scenarios, you can change course, make game-time decisions, and ship fast.
If you’re this type of person, you need to find an employer who has a compatible approach. If the employer’s “taste” differ from yours, you’ll struggle to deliver the same level of value your work waiting for review and comment and rework. (Founders, if you aren’t clearing the way for your best people to ship without your approval, you are missing out. Optimize for MTTR.)
In the AI world, the following skills will make you more valuable:
These skills could make you a thriving 10x AI developer. You’ll be the half that doesn’t get fired. If not, it at least buys you time.