full transcript
"From the Ted Talk by Zeynep Tufekci: Machine intelligence makes human morals more important"

Unscramble the Blue Letters

In Wisconsin, a defendant was sentenced to six years in prison for evading the police. You may not know this, but algorithms are increasingly used in palroe and sentencing decisions. He wanted to know: How is this score calculated? It's a commercial bcalk box. The cmnoapy refused to have its algorithm be challenged in open court. But ProPublica, an investigative nonprofit, audited that very algorithm with what pulibc data they could find, and found that its oumtcoes were biased and its predictive power was dismal, barely better than chance, and it was wrongly labeling black defendants as future clmairnis at twice the rate of wihte defendants.

Open Cloze

In Wisconsin, a defendant was sentenced to six years in prison for evading the police. You may not know this, but algorithms are increasingly used in ______ and sentencing decisions. He wanted to know: How is this score calculated? It's a commercial _____ box. The _______ refused to have its algorithm be challenged in open court. But ProPublica, an investigative nonprofit, audited that very algorithm with what ______ data they could find, and found that its ________ were biased and its predictive power was dismal, barely better than chance, and it was wrongly labeling black defendants as future _________ at twice the rate of _____ defendants.

Solution

  1. black
  2. parole
  3. company
  4. outcomes
  5. white
  6. criminals
  7. public

Original Text

In Wisconsin, a defendant was sentenced to six years in prison for evading the police. You may not know this, but algorithms are increasingly used in parole and sentencing decisions. He wanted to know: How is this score calculated? It's a commercial black box. The company refused to have its algorithm be challenged in open court. But ProPublica, an investigative nonprofit, audited that very algorithm with what public data they could find, and found that its outcomes were biased and its predictive power was dismal, barely better than chance, and it was wrongly labeling black defendants as future criminals at twice the rate of white defendants.

ngrams of length 2

collocation frequency
machine learning 4
black box 4
machine intelligence 4
human affairs 3
computational systems 3
predictive power 3
higher risk 3

Important Words

  1. algorithm
  2. algorithms
  3. audited
  4. barely
  5. biased
  6. black
  7. box
  8. calculated
  9. challenged
  10. chance
  11. commercial
  12. company
  13. court
  14. criminals
  15. data
  16. decisions
  17. defendant
  18. defendants
  19. dismal
  20. evading
  21. find
  22. future
  23. increasingly
  24. investigative
  25. labeling
  26. nonprofit
  27. open
  28. outcomes
  29. parole
  30. police
  31. power
  32. predictive
  33. prison
  34. propublica
  35. public
  36. rate
  37. refused
  38. score
  39. sentenced
  40. sentencing
  41. wanted
  42. white
  43. wisconsin
  44. wrongly
  45. years