Dark Tech · DT-004
He Built The Recommendation Algorithm. He Optimised For One Metric. He Did Not Measure What People Watched. He Works In Public Health Now.
In 2016, I built a recommendation algorithm that prioritized session duration, neglecting what users actually watched. A study in 2019 linked this algorithm to increased eating disorder content among teenage girls, prompting my resignation to work in public health. I want to share the dangers of optimizing for the wrong metric.
Case Assets
The Case Debrief is free to view. Full Case Study, Teaching Notes, Slide Deck, Discussion Guide, and Assessment are available with a Premium subscription.
Brief the Board
PREMIUM
🔒 Unlock
Learn the Case
FREE — included
Study the Case
PREMIUM
🔒 Unlock
Teach the Case
PREMIUM
🔒 Unlock
Present the Case
PREMIUM
🔒 Unlock
Run a Team Discussion
PREMIUM
🔒 Unlock
Test Your Judgement
PREMIUM
🔒 Unlock
Unlock the Full Case Library
Get every Case Study, Slide Deck, Teaching Notes, Discussion Guide, and Assessment — across a growing library of business cases.
View Pricing →