This question is on everyone’s mind. Sometimes it’s asked openly, other times it’s the elephant in the room. It’s clear that people sense a shift. Have software developers disrupted themselves out of a job?

For the past six months, I’ve been coding with AI tools almost daily. Here’s what I’ve learned:

Can you hit a moving target from a moving vehicle. That’s software development in a nutshell. The target (user demands, competitive pressures, market forces) constantly shifts. The vehicle (development environment) is piloted by multiple developers, not just one.

Now imagine giving the wheel to a superintelligent monkey. One that has a tendency to occasionally lie or behave erratically. That’s AI-generated code in our current development ecosystem. You can guess how well that goes.

Software development, at its core, involves modeling the real world—a task that’s messy and not easily described in text. While AI excels at processing textual information, humans learn through all their senses and interactions with others. This fundamental difference means that new feature development will likely remain in human hands for the foreseeable future.

This reality is either comforting or frustrating, depending on where you stand in the software development ecosystem. But it’s not the complete picture. I still believe software development will undergo multiple step function changes in a short period – without developers losing their jobs.

Here are three areas where you can start today and complete 90% of the work with minimal user intervention.

Legacy Code Migration: Rewrite large parts of your codebase to your preferred language or migrate to a modern framework at an incredible speed.

Writing Tests: AI can generate 90% of your unit and integration tests with minimal human intervention.

Self-Serve Codebase: Static documentation will become obsolete. Anyone can query the codebase in plain language. Imagine a sales engineer on a call answering a customer’s question about an obscure configuration option by simply querying the code.

Integrations, client libraries, enterprise software customization are other examples of development work that can be accelerated significantly using AI code generators.