Avoid Smooth Talking at Your Own Peril

The Unreasonable Effectiveness of Signaling Competence

Surely you’re not just fucking with me, Mr. Feynman

Richard Feynman was already semi-famous when some technicians pulled him into a room to review blueprints for a plant that hadn’t been built yet. The blueprints were bigger than the table. He had no idea what he was looking at.

So the story goes, he picks a spot on the diagram, points at it, and asks “what’s that?”

The engineers traced through the blueprint, working through the implications. Then they turned around, and said: “You’re absolutely right, sir.” Then they rolled up the blueprints and left.

Feynman’s military escort was amazed. How did he know?

He didn’t. He’d pointed at something at random and asked a question vague enough that the engineers figured out the problem themselves and then credited him with spotting it.

Feynman is the archetypal technical master. But he was also, apparently, a master at signaling competence.

The two skills often travel together in very successful people. But when they come apart, it’s rarely the technical skill that matters most.

Hard Skills vs Smooth Talk

Technically skilled people often think that smooth talking is someone else’s problem.

If you’re good enough at your thing, you will be too good to be ignored. But this is often wrong.

Smooth talking isn’t just fluff. In non-technical arenas, it is what people use to determine how useful you are to them.

This matters because signalers often control resources, write the checks and decide on promotions.

Signalers reward other signalers because they are more legible and more often lead to their own goals being achieved.

So, if you care about things like controlling resources, signalling is very much your problem.

Tug of war

A founder I respected learned this the hard way.

Being an obsessive skill-builder, he happily delegated all his fundraising to his cofounder because he found it boring.

His cofounder was a kind of idiot-savant, really good at smooth talking, but not particularly good at any other aspect of running a business.

My buddy ran this business top to bottom, in blissful isolation from the investors.

When they eventually had to go their separate ways, the investors rallied around the cofounder and pushed him out.

He was surprised by this.

But the investors weren’t stupid, they knew who was running the show. Their return just relied on someone raising another round in twelve months, and the cofounder was the best draft pick for that job.

My friend got wrecked because he failed to adapt to an environment whose incentives had gradually shifted from demonstrating competency to signaling it.

Rookie mistake.

Scaling Up

The higher up you go, the more the game shifts towards signaling skill.

If you’re the college-grad building the antenna on the Mars rover, your technical chops probably matter most.

But in the rooms with people deciding if it’s worth scaping dust off Mars in the first place, you’ll find more people who are signalers first and engineers second.

The shift makes sense.

Lower-level goals are concrete: optimize this code to run 12% faster. Higher-level goals are subjective: be a good manager, build a good culture, deliver a compelling pitch.

The more removed you are from object-level work, the more you’re being evaluated on your ability to get others to do things.

So the ratio shifts, continuously, with every step up.

Git Gud

The most effective people have technical competency AND signaling skill.

Feynman was a great physicist and, when he needed to be, a really good signaler.

When I explain this to technically-minded people, I can feel them flinch. It’s uncomfortable to think that your pure craft is muddied by simple monkeys jockeying for status.

When I explain this to people pitching visions for a living, they tend to confidently insist that they are, in fact, very good at doing things.

When I talk about this to people who have both skills, they just say “duh” and go back to taking over the world.