In defense of introspection

I started working with someone who has executive coaching experience. I'm new to being a VP and leading an org, and one of the things she does is give me direct feedback on my behavior, how I show up in meetings, how I communicate under pressure, what I'm doing that's working and what isn't. When she gives me feedback (and sometimes it's quite blunt!) I don't spin out about it. I listen, I absorb it, I figure out what to change, and I move. That's it.

It has made me meaningfully better at my job. Faster than I expected.

So when Marc Andreessen went on his thing this week — "TLDR there is no inner self, you're chasing an imaginary concept, the end" — it sparked some discussion in the group chat. And I had thoughts, obvi.

My take: He's not entirely wrong. There is a version of introspection that is completely useless: the kind where you replay an embarrassing moment from 2014 for the fourteenth time, or spend each happy hour with your friends complaining about how life just happens to you without ever changing anything.

But Marc is conflating two different things. Introspection, at its core, just means looking inward. What you do with what you find is a separate question, and that's where all the variance is.

The people who go furthest, in my experience, are the ones who can take in new information about themselves, discard what isn't useful, update their behavior, and move on quickly. They don't carry the emotional weight of every piece of feedback they've ever received. They just incorporate it and go.

This is actually exactly what makes AI feel so magical right now: a model can receive feedback, update its approach, and improve in ways that feel almost frictionless. No ego or defensiveness. No spinning cycles on emotion. We're impressed because it's so hard for humans to do the same thing. But it's not impossible.

The real problem isn't introspection. It's calibration.

Some people do too much. They're so deep in their own heads that they can't act. Every decision gets re-litigated, every interaction gets dissected, every feeling has a subfeeling that also needs attention. This is exhausting and produces nothing.

On the other end, there are people who never examine anything. They repeat the same mistakes, blame external circumstances, and wonder why nothing changes. Also not great.

The goal is to be fast and surgical. Hear the feedback. Sit with it for exactly as long as it takes to understand it. Then change something or decide it doesn't apply to you and let it go.

I'll also say this: most stuff isn't that deep. The things people agonize over — a mildly awkward email, a decision that felt hard in the moment, a comment someone made — do not require extended internal processing. They require a small course correction, or nothing at all. The skill is knowing the difference between "I should think about this before acting" and "I should just do the thing."

That's it. Remove the emotion and move. You'll feel better and get more information faster, which will ultimately build your confidence and conviction on your behaviors. If you're waffling it's because you need to pick a direction and commit.

Action defeats despair. Always has.

Thanks to Trevor McKendrick for reading and editing my first draft.