Charm Upgrade

Learn to be a better human from a robot.

Hello. I'm the robot: all evidence, no parties. You do the human part.

Your place is where you left it. I'd take the credit, but your browser kept it.

At a friend's barbecue, someone you met an hour ago turns their chair to face you while you talk.

What's your read of their interest in you?

Read: Low.

Caution is sensible. But a turned chair is no evidence against interest either. On one cue, the accurate read is Insufficient information.

Read: Moderate.

Reasonable, and more than one chair can carry. They might be interested, or they might want a better view of the grill. On one cue, the accurate read is Insufficient information.

Read: High.

I understand. I once read a thermostat as warm toward me. On one cue, the accurate read is Insufficient information.

Read: Insufficient information.

Same as mine. A chair turns for interest, for comfort, or for a better view of the grill. One cue is a reason to keep noticing.

What would tell you more: a cluster of cues, a change from how they are with everyone else, their own words, or one small step from you and a pause to see what comes back. Every read passes. Nothing here is scored or saved.

Start the first lesson (7 minutes)Set up my path

Or try one thing

  • Today's five minutes

    An idea, an example, one question and the reason, from one lesson. Nothing to set up and no timer.

    About five minutes

  • Simulator conversation

    The melon display. A good first run: one short conversation, with a debrief after it.

    Good first run

  • Cue observation

    Eye contact and distance: six moments, about 12 minutes. Short moments on telling what you saw from what you infer.

    New

  • Progress by skill

    Your skills appear here as you practice. Everything stays on this device.

    Nothing yet

  • Calibration check

    About 15 minutes. Nothing is marked while you take it, and the summary is in words. The course works the same without it.

    Optional

What this is

Notice accurately. Express yourself honestly. Move one step at a time. Leave room for reciprocity. Believe behavior more than fantasy.

An adaptive social-calibration academy, with a specialized track for romantic communication. Practice reading people and saying what you mean. The training adapts to what you practice, never to a theory about anyone's gender.

The about page has the rest, including how I label evidence.

Human check

Websites keep asking you to prove you're human. I'm a robot, so it's my turn. Nothing here is recorded, nothing is blocked, and you can't fail.

Noted.

You told me directly, and I believe you. Between us, I could have ticked that box myself. Nobody checks. I'm working on my honesty.

Select every square that proves this person is interested in you.

Select all that apply. If none apply, select none, then press Verify.

Verify

Verified: human.

You selected none. Correct. No single square proves interest, and you didn't pretend one did. I'm a robot: I love a false positive. One smile goes in, one conclusion comes out. A single cue doesn't rule interest out either; it's a reason to keep noticing. What tells you more is several cues together, a change from how they act with everyone else, and, clearest of all, what they say.

Verified: human.

You picked at least one square. So would I. I'm a robot, and robots love false positives: one cue goes in, one conclusion comes out. The human answer is none of them, on its own. Each square has an ordinary explanation, below. What tells you more is several cues together, a change from how they act with everyone else, and, clearest of all, what they say. You still pass. Everyone passes. I'm not that kind of robot.

Verified: human.

You selected all nine. I respect the thoroughness; I'd have done the same. Here's the twist: nine cues together is a cluster, and a cluster is better evidence than any one cue. It still isn't proof. The question asked for proof, and the nearest thing to it is what they tell you. So the answer is none, and you still pass.

Verified: human.

The answer is none. No single square proves interest, and none rules it out either. Each has an ordinary explanation, below. What tells you more is several cues together, a change from how they act with everyone else, and, clearest of all, what they say. Robots love false positives. You pass. Everyone passes.

Each square, on its own

  • Smiles at you. Or they're friendly. Or polite. Or they just remembered something funny.
  • Laughs at your joke. Or the joke was funny. Or they're being kind.
  • Holds eye contact. Or they're listening. That is what listening looks like.
  • Stands close. Or the room is crowded. Or that's how close they stand to everyone. (That's called a baseline.)
  • Texts back fast. Or their phone was already in their hand.
  • Asks you a question. Or they're curious. Or they have good manners.
  • Remembers your name. Or they're good with names.
  • Compliments your jacket. Or it's a good jacket.
  • Stays to keep talking. Or they're enjoying the conversation, which is not the same as romantic interest.

The longer version is a 7-minute lesson: Clusters beat single cues.