How I Contaminated 500 Experiments (By Talking Too Much)

Dear Magicians,

Feynman taught: Don’t fool yourself.

Turns out, I’d been fooling myself for 5 years—every time I filled a silence I should have measured.

Let me explain.

I run a podcast. You may have heard of it. And if I forget to mention it, the trusty folks at Reddit are sure to remind me how bad a podcaster I am.

Here’s what I’ve learned after 500+ episodes: I talk too much.

Shocking, I know. A professor who loves the sound of his own voice.

But here’s the thing. Every experiment needs a detector. For years, I thought my detector was my mouth. I’d ask a question, get three seconds of silence, and immediately start explaining what I thought the answer should be.

You know what I was actually doing? I was contaminating the experiment. Like shining a flashlight directly into a telescope and wondering why I couldn’t see any stars.

So recently I tried something radical. I started shutting up.

Not forever. I still have a podcast to fill. But I began treating silence the way a physicist treats a vacuum chamber—as something precious that shouldn’t be polluted.

The results were immediate.

And humbling.

When I stopped filling every pause with my own brilliant thoughts, I discovered something uncomfortable: most of my questions weren’t actually questions. They were conclusions with a question mark stapled to the end.

“Don’t you think that...” isn’t a question. It’s a sales pitch wearing a disguise.

Real questions require actual silence. The kind that feels awkward. The kind that makes you want to jump in and save everyone from the discomfort.

Think about it. When you’re talking, you already know what you’re saying. The information content, for you, is exactly zero. You’re just broadcasting something that’s already in your head.

But when you’re silent?

That’s when you can learn something new.

I started tracking this. After reading brutal reviews of a dozen podcasts last year, I had real data. The conversations where I spoke 40% of the time were twice as highly rated as the ones where I spoke 60% of the time.

It’s tricky because I often am a guest on other people’s podcasts. But this is no excuse. Mathematics doesn’t fib. But apparently I do to myself if not you dear listener—every time I pretend my verbal diarrhea is adding value.

Now, I’m not suggesting you take a vow of silence. That’s a different kind of fooling yourself. I tried that once at a faculty meeting. People thought I was having a stroke.

What I’m suggesting is this: Treat silence like the scientific instrument it actually is.

Silence isn’t the absence of signal. It’s the highest-resolution detector you’ll ever build.

Your colleague asks you a question? Pause. Actually think. Let the silence do some work.

I know what you’re thinking. “But silence is uncomfortable!”

Yes. So is observing the CMB from the South Pole. We do it anyway because the data is worth it.

The discomfort of silence is just your ego screaming for attention. It wants to fill the space with anything—even garbage—just to prove you’re smart.

So this week, try these small experiments, (partially inspired by this article)

Pause 5 seconds before responding to a question

Catch yourself when you start a sentence with “Don’t you think...”

Track one meeting for your talk-time ratio

Ask a genuine question and actually wait for the answer.

Over time, you might notice:

When you’re “performing” vs. “detecting”

That quiet colleagues aren’t passive—they’re just contemplative.

What might get in your way:

Silence is genuinely uncomfortable (social mammal wiring)

Academic culture actively rewards verbal performance

Until next time, have a M.A.G.I.C. Week,

Brian