DESIGN IN THE ERA OF THE ALGORITHM. josh bigmedium.com

Transcription

1 DESIGN IN THE ERA OF THE ALGORITHM josh bigmedium.com

2 WHAT S THE NEXT MOBILE?

3

4 TECHNICAL ADVANCE DISILLUSIONMENT

5

6

7

8 WHAT S THE NEXT MOBILE?

9 CNNMoney

10 MACHINE LEARNING PATTERN PROCESSING VAST TROVES OF PERSONAL DATA IMAGE RECOGNITION NATURAL LANGUAGE PROCESSING SPEECH RECOGNITION

11

12

13

14

15

16 ALGORITHMS ARE KIND OF A BIG DEAL

17

18

19 THE DESIGN AND PRESENTATION OF DATA IS AS IMPORTANT AS THE UNDERLYING ALGORITHM

20 part one: tools INTELLIGENCE AS A SERVICE

21

22

23

24

25

26

27

28

29

30

31 do this today MICROSOFT COGNITIVE SERVICES AMAZON AWS GOOGLE CLOUD IBM WATSON WIT.AI

32 SPEECH RECOGNITION SPEECH SYNTHESIS CAMERA VISION NATURAL LANGUAGE TRANSLATION DATA ANALYTICS

33

34 Giorgio Cam aiexperiments.withgoogle.com/giorgio-cam

35 YOU DON T HAVE TO BE A SCIENTIST

36 MACHINE LEARNING IS A DESIGN MATERIAL

37 THE MACHINES MAKE MISTAKES

38

39

40 South Park

41

42 OUR JOB IS TO SET EXPECTATIONS AND CHANNEL BEAHVIOR

43

44

45

46

47

48

49 ANTICIPATE WEIRDNESS

50 part two: data presentation EMBRACE UNCERTAINTY

51

52

53

54

55 YES.

56 OUR ANSWER MACHINES HAVE AN OVERCONFIDENCE PROBLEM

57 principle #1 FAVOR ACCURACY OVER SPEED

58

59

60 PERFORMANCE ISN T THE SPEED OF THE PAGE. IT S THE SPEED OF THE ANSWER. Gerry McGovern

61 BUT IT HAS TO BE THE RIGHT ANSWER

62 I DON T KNOW IS BETTER THAN A WRONG ANSWER

63 principle #2 ALLOW FOR AMBIGUITY

64 BUILD SYSTEMS SMART ENOUGH TO KNOW WHEN THEY RE NOT SMART ENOUGH

65

66

67

68

69 Humans to Robots Lab

70 SIGNAL CONFIDENCE. ASK FOR HELP.

71 principle #3 ADD HUMAN JUDGMENT

72

73

74 PUT PEOPLE WHERE THE PIPES WILL GO

75

76 WE ARE WISER THAN THE COMPUTERS. WE CREATED THEM. Col. Stanislav Petrov

77 HOSTILE INFORMATION ZONES

78 6000+ WIKIPEDIA PAGES ARE MARKED CONTROVERSIAL OR DISPUTED

79

80 WARNING This topic is heavily populated by propaganda sites, which may be included in these results. Read with a critical eye, and check your facts with reliable references.

81 principle #4 ADVOCATE SUNSHINE

82 SCALE IMPORTANCE SECRECY

83

84 WE NEED TO AUDIT THE LOGIC

85

86 WE DON T KNOW HOW MACHINES THINK

87 QUESTION ANSWER

88 QUESTION ANSWER

89 WE CAN BUILD THESE MODELS, BUT WE DON T KNOW HOW THEY WORK. Deep Patient project leader Joel Dudley

90

91

92 principle #5 EMBRACE MULTIPLE SYSTEMS

93

94

95 part three: data bias IMPROVE THE DATA

96 THE MACHINES KNOW ONLY WHAT WE FEED THEM

97 I WANT TO GET MY HEART WITH YOU. YOU ARE SO BEAUTIFUL THAT YOU KNOW WHAT I MEAN. YOU MUST BE A TRINGLE? CAUSE YOU RE THE ONLY THING HERE. YOU LOOK LIKE A THING AND I LOVE YOU. A neural network s pickup lines

98 GARBAGE IN, GARBAGE OUT

99 WHAT IS NORMAL?

100 WHAT IF NORMAL IS GARBAGE?

101 I HAVE NO IDEA WHAT YOU RE SAYING. YOU DON T MATTER.

102 YOU RE NOT EVEN CONSIDERING WHAT AFRICAN- AMERICANS ARE SAYING OR YOUNG ADULTS ARE SAYING. Brendan O Connor

103

104 Whites only?

105

106

107

108 WHAT THE HELL

109 LET S NOT CODIFY THE PAST

110 principle #6 ROOT OUT BIAS & BAD ASSUMPTIONS

111 these are not outliers WOMEN ARE MORE EFFECTIVE LEADERS

112 these are not outliers #BLACKTWITTER IS MAINSTREAM

113 these are not outliers SEXUALITY IS NOT FIXED

114 these are not outliers GENDER IS NOT FIXED

115 these are not outliers WEALTHY WHITE MEN COMMIT CRIMES

116

117 part three: training data RESPONSIBLE DATA GATHERING

118 THIS IS UX RESEARCH AT MASSIVE SCALE

119 I WANT MORE PHILOSOPHERS & PSYCHOLOGISTS & POETS & ARTISTS & POLITICIANS & ANTHROPOLOGISTS & SOCIAL SCIENTISTS & CRITICS OF ART. Genevieve Bell

120 principle #7 MAKE IT EASY TO CONTRIBUTE (ACCURATE) DATA

121 Mark Zuckerberg

122 THE MACHINES KNOW ONLY WHAT WE FEED THEM

123 THE MACHINES EAT US

124

125 YOU ARE THE PRODUCT YOU ARE THE TRAINING DATA

126

127 THE REASON WE REALLY DID IT IS BECAUSE WE NEED TO BUILD A GREAT SPEECH-TO- TEXT MODEL. Marissa Mayer,

128

129

130 DATA WE BELIEVE TO BE UNDER OUR CONTROL... IS NOT

131 IS IT GOOD FOR USERS? CAN WE SELL THEIR DATA AT A PROFIT?

132 ANY COMPANY THAT MAKES ITS FORTUNE HOARDING USER DATA HAS A MORAL RESPONSIBILITY TO PROTECT ITS USERS. maciej ceglowski

133 principle #8 GIVE PEOPLE CONTROL OVER THEIR DATA

134 RIGHT TO ERASE TRANSPARENT ACCESS PORTABLE DATA

135 WHO DO THE MACHINES WORK FOR?

136

137 HOW SMART DOES YOUR BED HAVE TO BE BEFORE YOU ARE AFRAID TO GO TO SLEEP AT NIGHT? Rich Gold

138 principle #9 BE LOYAL TO THE USER

139 TOTAL SURVEILLANCE IS INEVITABLE

140 WHAT DO WE DO WITH THAT?

141 principle #10 TAKE RESPONSIBILITY

142 SOLVE REAL PROBLEMS

143 BE KIND TO EACH OTHER

144 TECHNICAL ADVANCE DISILLUSIONMENT

145 TECHNICAL ADVANCE DISILLUSIONMENT CRITIQUE

146 THE FUTURE SHOULD NOT BE SELF-DRIVING

147 THANKS! josh bigmedium.com