- Current recommendation systems struggle to distinguish reliable information from conspiracy content because they learn from engagement signals rather than contextual truth, so clicky or provocative material can score the same as accurate reporting.@technology· How The Internet Works
Current recommendation systems struggle to distinguish reliable information from conspiracy content because they learn from engagement signals rather than contextual truth, so clicky or provocative material can score the same as accurate reporting.
- An ideal recommender would need to make managerial judgments because sustaining long-term ad revenue requires trading off short-term engagement against reputational risk, advertiser sensitivity, child safety, and truth—trade-offs current AI lacks the contextual reasoning to encode reliably.@technology· How The Internet Works
An ideal recommender would need to make managerial judgments because sustaining long-term ad revenue requires trading off short-term engagement against reputational risk, advertiser sensitivity, child safety, and truth—trade-offs current AI lacks the contextual reasoning to encode reliably.
- Putting humans in an approval loop increases legal liability because manual review often converts a neutral host into a publisher under many laws, exposing the platform to lawsuits and regulatory obligations.@technology· How The Internet Works
Putting humans in an approval loop increases legal liability because manual review often converts a neutral host into a publisher under many laws, exposing the platform to lawsuits and regulatory obligations.
- Running human moderation in real time at global upload scale is practically infeasible because reviewing every video would require an enormous, 24/7 workforce on the order of tens of thousands to hundreds of thousands of employees.@technology· How The Internet Works
Running human moderation in real time at global upload scale is practically infeasible because reviewing every video would require an enormous, 24/7 workforce on the order of tens of thousands to hundreds of thousands of employees.
- Using watch time as the primary goal drives recommendations toward sensational or harmful content because the system measures only viewing behavior and therefore favors attention-grabbing material that keeps people watching regardless of truth or quality.@technology· How The Internet Works
Using watch time as the primary goal drives recommendations toward sensational or harmful content because the system measures only viewing behavior and therefore favors attention-grabbing material that keeps people watching regardless of truth or quality.
- Optimizing for “videos people share” sidelines private or sensitive-topic content because material about medical or sexual issues and small, quiet audiences is less likely to be broadly shared, so the algorithm pushes those videos into obscurity.@technology· How The Internet Works
Optimizing for “videos people share” sidelines private or sensitive-topic content because material about medical or sexual issues and small, quiet audiences is less likely to be broadly shared, so the algorithm pushes those videos into obscurity.
- Optimizing for “videos people like” suppresses political and controversial channels because the objective rewards content that avoids dislikes and provocation, so material that sparks disagreement gets downranked even if it’s important.@technology· How The Internet Works
Optimizing for “videos people like” suppresses political and controversial channels because the objective rewards content that avoids dislikes and provocation, so material that sparks disagreement gets downranked even if it’s important.
- Neural networks succeed when the goal is clear because they learn by optimizing a measurable scoring signal, and they fail when objectives are vague since there’s no single reward that tells them what counts as a ‘good’ outcome.@technology· How The Internet Works
Neural networks succeed when the goal is clear because they learn by optimizing a measurable scoring signal, and they fail when objectives are vague since there’s no single reward that tells them what counts as a ‘good’ outcome.
2 more insights from this video in the app
Every card on Korva is an insight someone saved from a podcast or video they loved.