In the Long Run, Everything is a Fad

The Olympic Gymnastics Disaster

The 2024 Paris Olympics floor exercise final became a 457-day legal nightmare:

1. Jordan Chiles scored 5th, but her coach noticed a scoring error
2. Judges agreed → she moved to 3rd (bronze medal)
3. Anna Barbosu (Romania) filed complaint: the appeal was 1 minute 4 seconds late (limit: 1 minute)
4. Court of Arbitration ruled: Jordan's score reverted, Anna gets bronze
5. Sabrina Voinea noticed: her out-of-bounds penalty was wrong (video shows she was in bounds)
6. More lawsuits. Swiss Federal Supreme Court. Still unresolved.

The lesson: We built elaborate quantification systems (the "Code of Points") to replace corrupt judges—and got something even more broken.

The Generational Cycle

Research shows people believe the best music, fashion, economy, and morals were during their teenage years.

For our generation in data: The "best stuff professionally" was Moneyball, Nate Silver, rigorous quantification. We rebelled against pundit vibes.

But there's always a next generation...

The Rise of Vibes

Evidence it's already happening:

• Zohran Mamdani won NYC mayoral primary on TikTok vibes, not data-driven campaigns
• "Taste is eating Silicon Valley" — products built on craft, not optimization
• Dating apps now match via AI reading profiles, not algorithmic compatibility scores

The Fujitsu Gymnastics AI

Page 2 of the Olympic Code of Points? An ad for Fujitsu's 3D-sensing AI scoring system.

It doesn't:

• Parse the code of points
• Calculate D-scores and E-scores
• Apply penalty deductions

It does:

• Watch the routine
• Compare to millions of examples
• Give a vibe-based score

It's the perfect judge—not because it's more rigorous, but because it's seen every gymnastics routine ever.

The Dirty Secret of Quantification

That objective-looking gymnastics scoring system? Full of subjective decisions:

• Why 8 elements, not 10?
• Why these point values?
• What is "poor rhythm"?
• Why can't some scores be reviewed?

"The best stuff professionally isn't math. It's numeric vibes."

The Business Reality

Boss: "Is our business in good shape?"

Analyst: "Well, it depends on what you mean by 'good'..."

Boss: 😐

We give answers like "What's an active user? We have conservative and aggressive definitions..." when they want a straight answer.

But what if there's a tool that just reads all the support tickets and says: "Customers are frustrated about X, Y, Z"?

The Faith Problem

"There's a ton of faith in data work. How do we know it works? We hire more data people—they'll tell us."

This faith is fragile. If:

• Quantification leads to 457-day lawsuits
• Tools don't really work
• Everything else runs on vibes (code, dating, politics)
• And AI vibes actually work...

Then we become the old geysers complaining that things were better in our day.