For most of human history, you could learn two things about a stranger in about ten seconds.

If they were lean and muscular, they were disciplined. Genetics helped, but nobody got that look without years of saying no to dessert and yes to the gym or hard work. If they wrote or spoke clearly and laid out an argument that held together, they were smart. You couldn't fake either one. The signal was expensive to produce, which is exactly what made it worth trusting.

Both signals are breaking at the same time, and I don't think most people have noticed.

GLP-1 drugs have turned down the appetite dial for millions of people who spent decades losing the willpower fight. Hormone replacement is now a routine conversation for men and women over 40. And the next wave targets muscle directly: Lilly paid roughly $2 billion for bimagrumab, an antibody that in trials adds lean mass and drops fat with no workout involved. It's still experimental, but the direction is obvious. Within a decade, the lean, muscled look that used to take years of discipline will be something you pick up at the pharmacy.

I say this as someone who has spent years treating my body like an engineering project: tracking biomarkers, lifting, watching what I eat (check out http://bionic.robbieallen.com). I'm not against any of these drugs. Some of them are the best thing to happen to metabolic health in my lifetime. But I've noticed something. The look I worked for no longer tells anyone what I did to get it.

Then there's intellect. I've been building software that writes for more than 15 years. At Automated Insights we were doing generative AI before anyone called it that, so I've watched this one coming for a long time.

A crisp email, a tight memo, a clever LinkedIn post: these used to be proof that someone smart had done the work. Now they're evidence of a $20 subscription. The output looks identical. The thing it used to prove is gone.

Case in point: this piece. I dictated the idea into my phone this weekend, rambling and half-formed. An AI turned it into a first draft. I rewrote it. The ideas are mine. The typing mostly wasn't.

Biologists call these costly signals. A peacock's tail works because a sick peacock can't grow one. The moment a signal gets cheap, it stops carrying information, not because anyone is lying but because the correlation is gone. A six-pack used to mean discipline. Now it means a prescription. Polished prose used to mean intelligence. Now it means access.

We're about to live in a world where the two most common shortcuts for judging people are worthless. So what do you look at instead?

Disclosure becomes the currency. The problem was never using the tool. Presidents have used speechwriters for a century and nobody calls them frauds. The problem is hiding it. People who say "here's what's me and here's what's augmented" will earn more trust than people with perfect output and no explanation. That's why I told you how this piece got written.

Outputs over proxies. I see a lot of pitch decks at the NC Tweener Fund. They're all beautiful now, because beautiful is free. What a deck can't tell me is how a founder thinks when the question isn't one they prepped for. Same for hiring: the polished résumé tells you nothing. The interview can be gamed. The two-week trial project can't.

The stuff that's still expensive. Judgment under pressure. Taste. Showing up for ten years and being the same person the whole time. A live, unscripted conversation. These are the new costly signals, and you can't get any of them at a pharmacy or from a prompt box.

If you're running an AI rollout, this is already happening in your building. The status report is polished. The customer email is flawless. The analysis deck has a chart on every page. None of it tells you whether the person who sent it understands what's in it.

Last week I wrote about a CEO whose rule for his executives is simple: use as much AI as you want, but if you can't explain the answer, you don't belong in the meeting. That's a costly signal test. Walking through your own logic, live, when someone asks a follow-up question, is still expensive.

So is follow-through. I spent a month tracking every commitment people made on my calls. Roughly 20% got done. Saying "I'll send you that" costs nothing. Sending it is the signal.

And so is showing up. A business owner told me his reps get a warmer reception walking into buildings now than they did a few years ago, "because oh my god, you're a real human being." When every inbox is full of perfect AI outreach, a person in the doorway is the expensive thing.

Here's my conclusion, and it's not the one I expected. Maybe this is fine. We spent a century treating leanness as a proxy for virtue and articulateness as a proxy for intelligence, and both correlations were always weaker than we pretended. Plenty of disciplined people carry extra weight because of genetics. Plenty of brilliant people write badly. Decoupling the signal from the thing it was supposed to signal ends a status game that a lot of people lost through no fault of their own.

One caveat. AI-assisted output isn't the same as AI-substituted judgment. If the tool made you clearer, great. If the tool is doing your thinking, that's a different problem, and it shows up the first time someone asks a follow-up question.

We adapted to makeup, hair dye, ghostwriters, Photoshop, and Instagram filters. We'll adapt to this too. It just means unlearning the shortcuts and going back to the slower thing: paying attention to what people actually do.

What you see is no longer what you get. What's the signal you're still trusting that you probably shouldn't?

Two things I'd genuinely like back from you. First, hit Reply and tell me which part of this matches what you're seeing. Second, if there's someone in your world who should be on this list, forward it to them.

- Robbie