Human Judgement

I watched an excellent 45 seconds by Nate B Jones (here) which resonated with me as it aligns with how we are thinking about AI in our workflows.

AI is changing software engineering, but its not the one-dimensional impact as some people might think.

Nate frames this really well - it helps to think about this in three layers.

First, there is the work of generating artifacts. This includes specs, code, tests, docs, plans, summaries, analyses, and POCs. This is the part of the job where AI is having the biggest impact. The cost of producing a first pass has fallen dramatically and with well-thought intent and context AI produces good output.

Second, there is judgment and accountability. Someone still has to decide whether the generated output is correct, safe, maintainable, aligned with our standards, and appropriate for the business. Someone still has to own the decision that work is ready ship.

Third, there is execution in the real world. In software, that means operating production systems, managing risk, handling incidents, working through dependencies, and delivering outcomes under real constraints - something AI doesn't do well (today) and requires a human level of autonomy AI doesn't currently possess.

AI is strongest in the first layer. It is increasingly helpful in the second. It does not remove the third.

So the real shift is not that engineers become unnecessary. It is that generation becomes cheaper, while judgment, system understanding, and ownership become even more valuable.

That is the mindset we should have as engineering teams. We use AI to accelerate but do not confuse generated output with completed engineering.

AI can produce options. Engineers are still responsible for choosing the right one and making it work well in practice.