The most dangerous thing in AI right now is being “Too Experienced”

People often ask me how to stay relevant in AI, and I usually think back to a moment early in my career when I realised something uncomfortable: the people who stagnated the most weren’t the ones who knew too little, quite the opposite: they were the ones who were too attached to what had made them successful before.

If you want to survive (and honestly, thrive) in AI right now, you need a slightly dangerous mindset — you need to assume that more is possible than you are currently able to prove, and you need to be willing to explore that space before you feel fully ready, because if you only move once something is obvious and well-documented, you are already late.

Here is the part that is controversial when I say it to experienced engineers and data scientists: expertise can quietly become a comfort zone. Not because expertise is bad. Expertise is what keeps systems from breaking, keeps customers safe, and keeps teams from reinventing catastrophic mistakes. However expertise also builds very strong internal filters about what is “realistic”, “worth it”, or “not how things are done”.

The best people I see in AI today use expertise like guardrails on a mountain road: it keeps them from driving off the cliff, but it doesn’t decide where they are allowed to go.

Everything you know today is, by definition, about yesterday’s constraints. And this field is moving fast enough that yesterday’s constraints are often just historical artefacts.

So you have to actively practice intellectual lightness. Question your strongest opinions. Notice when you are defending an approach just because you invested years into mastering it. Be ruthless about dropping mental models that no longer create leverage, even if they once defined your professional identity.

Stay curious enough to explore things that feel slightly uncomfortable.

Stay humble enough to assume you are missing something important.

Stay ambitious enough to aim for impact, not just correctness.

Use what you know: be resourceful, be pragmatic, be responsible. But don’t become emotionally attached to it.

Because in this field, the real risk is not that you don’t know enough today.

The real risk is that you become the person who stops being able to learn what tomorrow requires.