Being junior in AI

Everyone tells you the field is moving fast. What they don't mention is that the people who've been doing this for twenty years are also figuring it out as they go.

That's the strange thing about being a junior designer in AI right now. The usual gap between junior and senior — the experience deficit, the pattern recognition you haven't built yet — is narrower than it's ever been in any design field I know of. Not because you know more. Because the field itself is so new that nobody has had time to accumulate the kind of knowledge that usually separates them.

This is either exciting or unsettling depending on the day. Sometimes both before lunch.

What experience doesn't cover

In most design work, you can lean on established patterns. There's a correct way to design a checkout flow. There's decades of research on navigation. You can be wrong, but the wrong answer is at least identifiable.

In AI product design, those anchors mostly don't exist yet. What does a good error state look like when the error is "the AI doesn't know"? How do you design onboarding for someone who has never used an AI agent and has no mental model for what it can do? How much should a user trust the output before acting on it?

These questions don't have settled answers. Senior designers don't have them either. Everyone is working from first principles, running experiments, watching what breaks.

What being new is actually useful for

The disadvantage of being junior is that you don't know what you don't know. The upside is the same thing looked at differently: you haven't learned to stop asking the obvious questions.

I ask a lot of obvious questions. Why does this flow work this way? What are we assuming the user already understands? What happens if they don't? Sometimes the answer is "we tried the alternatives and this held up better." Sometimes it's "good point, let's think about it."

The obvious question is underrated. Most design problems don't need more cleverness. They need someone willing to ask whether the thing is actually necessary at all.

The part nobody warns you about

Being junior in a fast-moving field is genuinely disorienting. The tools change fast. The best practices are still being invented. Most of what's written about AI design is either marketing or speculation, and telling the two apart takes time you don't always have.

What I keep coming back to: design the thing in front of you as carefully as you can. Talk to the people using it. Watch what confuses them. Change what doesn't work. That's not specific to AI. It's just how design works, and it turns out it's still true even when everything else is uncertain.

The field being new doesn't make the fundamentals less relevant. If anything it makes them more important, because they're the only solid ground there is.