Experiment 3

• Type 1 errors

  • Information overload for resource exhaustion
  • Deception very apparent, hard to miss
  • Hard to differentiate from real targets

• Results

  • Problem complex with deception not on
  • Differentiation slows attack with deception on
  • A cat and mouse game likely to result in escalation on all sides
  • Classic example of fooling machines thus forcing people to build better tools

Notes: