1
00:00:04,080 --> 00:00:07,480
Speaker 1: Welcome to Aaron Benky's Cabinet of Curiosities, a production of

2
00:00:07,480 --> 00:00:13,840
Speaker 1: I Heart Radio and Grim and Mild. Our world is

3
00:00:13,960 --> 00:00:17,960
Speaker 1: full of the unexplainable, and if history is an open book,

4
00:00:18,280 --> 00:00:21,640
Speaker 1: all of these amazing tales are right there on display,

5
00:00:22,000 --> 00:00:27,600
Speaker 1: just waiting for us to explore. Welcome to the Cabinet

6
00:00:27,840 --> 00:00:39,519
Speaker 1: of Curiosities. Everybody has to start somewhere. To become a

7
00:00:39,680 --> 00:00:42,640
Speaker 1: great leader or an expert in a particular field takes

8
00:00:42,720 --> 00:00:45,680
Speaker 1: years of experience. A person doesn't go from an entry

9
00:00:45,720 --> 00:00:49,280
Speaker 1: level position to the corner office overnight. One must work

10
00:00:49,320 --> 00:00:53,040
Speaker 1: their way up the ranks until they reach the top. Unfortunately,

11
00:00:53,200 --> 00:00:55,800
Speaker 1: some doors are closed so tight you can't even get

12
00:00:55,800 --> 00:00:58,560
Speaker 1: a toe in, let alone a whole foot. Sometimes you

13
00:00:58,560 --> 00:01:00,320
Speaker 1: have to take matters into your own hand. Ends to

14
00:01:00,320 --> 00:01:03,600
Speaker 1: get noticed, like James Dyson, who was so angry that

15
00:01:03,680 --> 00:01:06,399
Speaker 1: his vacuum kept breaking, he took it apart to see

16
00:01:06,400 --> 00:01:09,200
Speaker 1: how it worked. It turned out that a critical design

17
00:01:09,280 --> 00:01:11,640
Speaker 1: flaw was the cause of his problems, so he got

18
00:01:11,640 --> 00:01:14,520
Speaker 1: to work on a new kind of vacuum design, and

19
00:01:14,720 --> 00:01:17,880
Speaker 1: unlike the other models on the market, his didn't need

20
00:01:17,959 --> 00:01:22,200
Speaker 1: a bag. Dyson built over five thousand prototypes before finding

21
00:01:22,240 --> 00:01:26,000
Speaker 1: one that worked, doors were shut in his face, Manufacturers

22
00:01:26,040 --> 00:01:28,880
Speaker 1: refused to make it, and retailers wouldn't carry it in

23
00:01:28,880 --> 00:01:32,039
Speaker 1: their stores. It took a long long time before those

24
00:01:32,080 --> 00:01:36,200
Speaker 1: stores and their customers saw the benefits of a Dyson vacuum,

25
00:01:36,240 --> 00:01:40,480
Speaker 1: but once they did, the rest was history. Stewart Freeborn

26
00:01:40,640 --> 00:01:43,280
Speaker 1: was another person with a dream, one that he was

27
00:01:43,319 --> 00:01:45,600
Speaker 1: going to make come true, no matter what it took.

28
00:01:46,080 --> 00:01:50,320
Speaker 1: Born in London in nineteen fourteen, Freeborn was fascinated with makeup,

29
00:01:50,640 --> 00:01:53,880
Speaker 1: but not the fashion kind. The idea of dressing up

30
00:01:53,920 --> 00:01:58,040
Speaker 1: to look like someone or something else fascinated him, and

31
00:01:58,120 --> 00:02:00,560
Speaker 1: as he got older, he grew into did in the

32
00:02:00,640 --> 00:02:04,000
Speaker 1: chemistry of different plastics, learning how they were made and

33
00:02:04,040 --> 00:02:07,680
Speaker 1: how they affected the human skin when applied, and he

34
00:02:07,760 --> 00:02:10,400
Speaker 1: knew exactly how he wanted to put his experience to

35
00:02:10,560 --> 00:02:15,079
Speaker 1: use in the movies. Freeborn's father, an insurance broker, begged

36
00:02:15,120 --> 00:02:18,200
Speaker 1: him to join the family business, but Stewart couldn't shake

37
00:02:18,280 --> 00:02:21,720
Speaker 1: his love of movie magic. He even started sneaking onto

38
00:02:21,760 --> 00:02:24,760
Speaker 1: the lots of shows whenever he was able to, and

39
00:02:24,960 --> 00:02:28,320
Speaker 1: just like Dyson in his early days, no one was interested.

40
00:02:28,760 --> 00:02:32,639
Speaker 1: Refusing to give up, Freeborn made a bold move in Nive,

41
00:02:32,680 --> 00:02:35,359
Speaker 1: he called up the local newspapers and studios to let

42
00:02:35,360 --> 00:02:38,919
Speaker 1: them know that Highly Selassie, the Emperor of Ethiopia, had

43
00:02:39,000 --> 00:02:42,160
Speaker 1: arrived in London. Not long after, the Emperor was seen

44
00:02:42,280 --> 00:02:47,200
Speaker 1: driving around Freeborn's neighborhood, that is, until the police detained him.

45
00:02:47,280 --> 00:02:50,280
Speaker 1: Unfortunately for them, all they had caught was an ambitious

46
00:02:50,320 --> 00:02:53,440
Speaker 1: young makeup artist in a fake beard and nose prosthetic

47
00:02:53,480 --> 00:02:56,600
Speaker 1: looking for his big break, and his stunt didn't even

48
00:02:56,680 --> 00:02:59,560
Speaker 1: land him a job. It wasn't until a British director

49
00:02:59,639 --> 00:03:03,040
Speaker 1: named Alexander Korda took a chance on him, giving Freeborn

50
00:03:03,160 --> 00:03:05,800
Speaker 1: some work on period dramas being filmed at the time.

51
00:03:06,840 --> 00:03:08,920
Speaker 1: One of his first major productions was that of a

52
00:03:09,000 --> 00:03:13,000
Speaker 1: nineteen forty eight adaptation of Oliver Twist by Charles Dickens.

53
00:03:13,040 --> 00:03:15,799
Speaker 1: The role of Fagin would be played by Alec Guinness,

54
00:03:15,840 --> 00:03:19,040
Speaker 1: someone whom Freeborn would work with again on such films

55
00:03:19,080 --> 00:03:21,919
Speaker 1: as The Bridge Over the River Quai in nineteen fifty seven,

56
00:03:22,160 --> 00:03:24,919
Speaker 1: as well as a little science fiction film being shot

57
00:03:24,960 --> 00:03:28,720
Speaker 1: in the Tunessee and Deserts in nineteen seventy seven. In fact,

58
00:03:28,760 --> 00:03:32,480
Speaker 1: Freeborn was responsible for many of Star Wars colorful characters,

59
00:03:32,760 --> 00:03:37,120
Speaker 1: including one fan favorite. In particular, Concept artist Ralph mcquarie

60
00:03:37,280 --> 00:03:41,080
Speaker 1: had imagined Han Solo's loyal Wookie partner, Chewbacca, as a

61
00:03:41,160 --> 00:03:44,720
Speaker 1: seven foot tall monster with a misshapen head, yellow eyes,

62
00:03:45,000 --> 00:03:48,760
Speaker 1: sharp teeth, of vests and boots. Freeborn put his own

63
00:03:48,800 --> 00:03:51,920
Speaker 1: spin on the character, making him more doglike in appearance

64
00:03:52,160 --> 00:03:55,360
Speaker 1: and replacing his vest with the bandalier. The result was

65
00:03:55,440 --> 00:03:58,040
Speaker 1: a gentle giant with an iconic growl who would go

66
00:03:58,080 --> 00:04:02,680
Speaker 1: on to save the galaxy numerous times. The film's sequel, however,

67
00:04:02,880 --> 00:04:06,200
Speaker 1: had a unique problem. A new creature was needed, and

68
00:04:06,320 --> 00:04:09,760
Speaker 1: it too was required to look unique, something to set

69
00:04:09,760 --> 00:04:12,640
Speaker 1: it apart from the other characters that Freeborn had designed,

70
00:04:13,040 --> 00:04:15,600
Speaker 1: and so he turned to two unlikely places from which

71
00:04:15,640 --> 00:04:20,599
Speaker 1: to draw inspiration. A mirror at a photograph of Albert Einstein.

72
00:04:21,400 --> 00:04:24,800
Speaker 1: Upon seeing his own reflection, Freeborn noticed the topography of

73
00:04:24,880 --> 00:04:27,640
Speaker 1: his face, how the lines and bumps told the story.

74
00:04:28,120 --> 00:04:30,960
Speaker 1: He was older and more experienced than he had been

75
00:04:31,000 --> 00:04:33,640
Speaker 1: in his youth, and he lost so much of his hair.

76
00:04:34,360 --> 00:04:37,040
Speaker 1: As for Einstein, it was his eyes that had captivated

77
00:04:37,040 --> 00:04:39,200
Speaker 1: the artist the most. They were the eyes of a

78
00:04:39,320 --> 00:04:42,719
Speaker 1: quiet and wise man, and they were surrounded by Wrinkles.

79
00:04:43,160 --> 00:04:46,120
Speaker 1: Freeborn combined all the features the creases of his own

80
00:04:46,160 --> 00:04:49,960
Speaker 1: weathered face with Einstein's sunken eyes to create one of

81
00:04:50,000 --> 00:04:53,960
Speaker 1: the most recognizable characters in the entire Star Wars universe,

82
00:04:54,720 --> 00:05:11,039
Speaker 1: the two foot tall Jedi master Yoda. The funny thing

83
00:05:11,120 --> 00:05:14,360
Speaker 1: about technology is that not only is it constantly changing,

84
00:05:14,520 --> 00:05:17,839
Speaker 1: it's changing us as well. How we work and the

85
00:05:17,880 --> 00:05:20,920
Speaker 1: way we think are ever evolving thanks to Breakthrough is

86
00:05:20,960 --> 00:05:24,640
Speaker 1: happening every day. Telephones used to be bound to wires

87
00:05:24,640 --> 00:05:28,480
Speaker 1: in our homes and offices. As satellites and wireless technology

88
00:05:28,520 --> 00:05:32,159
Speaker 1: spread across the world, landlines gave way to cell phones. Today,

89
00:05:32,200 --> 00:05:34,479
Speaker 1: the devices we carry in our pockets are no longer

90
00:05:34,600 --> 00:05:38,240
Speaker 1: used only for voice communication. We record our lives using

91
00:05:38,279 --> 00:05:40,919
Speaker 1: their built in cameras. We can look up any fact

92
00:05:41,000 --> 00:05:43,880
Speaker 1: on the Internet, or post our every thought in two

93
00:05:43,960 --> 00:05:47,960
Speaker 1: hundred and forty characters or less. In other words, we've

94
00:05:48,000 --> 00:05:52,159
Speaker 1: come a long way today. We're the ones doing the changing.

95
00:05:52,520 --> 00:05:54,839
Speaker 1: The more we share about our lives, the more selfies

96
00:05:54,880 --> 00:05:57,320
Speaker 1: we post and ads we pick up on, the more

97
00:05:57,360 --> 00:06:01,440
Speaker 1: our technology learns about us, our behavior, years are buying habits,

98
00:06:01,480 --> 00:06:04,600
Speaker 1: even what we look like. This information can be used

99
00:06:04,600 --> 00:06:07,279
Speaker 1: to influence the things we purchase or make it easy

100
00:06:07,360 --> 00:06:10,960
Speaker 1: for governments to track us down. But for Hasashi Canby,

101
00:06:11,040 --> 00:06:15,680
Speaker 1: it made it easier to identify pastries. Cambay founded Brain

102
00:06:15,720 --> 00:06:20,520
Speaker 1: Company Ltd. In n after spending several years designing computerized

103
00:06:20,520 --> 00:06:23,760
Speaker 1: graphics systems. His work could be seen on nightly news

104
00:06:23,800 --> 00:06:28,440
Speaker 1: programs displaying baseball scores and monetary exchange rates. From there,

105
00:06:28,560 --> 00:06:32,560
Speaker 1: Camba moved on to digitizing the textile industry. He replaced

106
00:06:32,560 --> 00:06:35,160
Speaker 1: the handwritten card system that had been used to program

107
00:06:35,240 --> 00:06:39,000
Speaker 1: looms with an automated computer program. This resulted in fewer

108
00:06:39,040 --> 00:06:43,599
Speaker 1: mistakes and less wasted fabric. As the years progressed, Brain

109
00:06:43,720 --> 00:06:47,919
Speaker 1: continued to innovate with applications for bridge engineers, students, and

110
00:06:48,160 --> 00:06:51,280
Speaker 1: other kinds of textile manufacturers. But it was in two

111
00:06:51,360 --> 00:06:55,000
Speaker 1: thousand seven when everything changed. A restaurant chain was about

112
00:06:55,040 --> 00:06:57,599
Speaker 1: to venture into the baking business, and if it wanted

113
00:06:57,600 --> 00:07:01,680
Speaker 1: to compete, it had to offer a wide variety of pastries. Now,

114
00:07:01,720 --> 00:07:04,239
Speaker 1: in America, a bakery might sell a handful of different

115
00:07:04,320 --> 00:07:07,600
Speaker 1: kinds of cookies, cakes, and other goods. In Japan, though,

116
00:07:07,800 --> 00:07:12,400
Speaker 1: the most successful pastry shops sold hundreds of varieties. Unfortunately,

117
00:07:12,520 --> 00:07:16,480
Speaker 1: that success came with complication. Most items were sold without

118
00:07:16,480 --> 00:07:20,080
Speaker 1: wrappers or bar codes, meaning cashiers had to memorize the

119
00:07:20,120 --> 00:07:23,240
Speaker 1: look and price of each pastry. They also had to

120
00:07:23,280 --> 00:07:27,040
Speaker 1: handle each individual item, and this process led along lines

121
00:07:27,120 --> 00:07:30,160
Speaker 1: and contamination by the cashiers. There had to be a

122
00:07:30,200 --> 00:07:32,960
Speaker 1: better way, and the Brain Company was going to find it.

123
00:07:33,920 --> 00:07:36,640
Speaker 1: They got to work on an artificial intelligence program to

124
00:07:36,720 --> 00:07:40,560
Speaker 1: assist in identifying pastries on a conveyor belt. It used

125
00:07:40,640 --> 00:07:43,920
Speaker 1: multiple algorithms to detect the features of different kinds of

126
00:07:43,960 --> 00:07:48,000
Speaker 1: baked goods, something they referred to as computer vision, but

127
00:07:48,120 --> 00:07:51,160
Speaker 1: it was a tough task. For example, a croissant might

128
00:07:51,200 --> 00:07:54,480
Speaker 1: not always be crescent shaped, and some pastries looked just

129
00:07:54,600 --> 00:07:57,360
Speaker 1: like others in shape and size, the only difference being

130
00:07:57,400 --> 00:08:00,480
Speaker 1: their fillings or if they had sugar on top. Can

131
00:08:00,560 --> 00:08:03,080
Speaker 1: bas team kept working, though, even as the two thousand

132
00:08:03,160 --> 00:08:05,920
Speaker 1: eight financial crisis threatened to shut them down for good,

133
00:08:06,360 --> 00:08:09,640
Speaker 1: and as other projects lost funding due to budget cuts.

134
00:08:09,680 --> 00:08:12,960
Speaker 1: He poured one percent of his remaining resources into the

135
00:08:13,080 --> 00:08:18,000
Speaker 1: pastry project. Brain rewrote its algorithms, weaving them together with

136
00:08:18,080 --> 00:08:22,280
Speaker 1: complicated camera systems and lights. They also accounted for variations

137
00:08:22,280 --> 00:08:24,680
Speaker 1: like if a doughnut got split in half on the

138
00:08:24,720 --> 00:08:27,720
Speaker 1: conveyor belt, with the program know that two pieces were

139
00:08:27,760 --> 00:08:30,640
Speaker 1: meant to go together, or if multiple items were touching,

140
00:08:30,680 --> 00:08:33,840
Speaker 1: would it understand that they weren't. One giant pastry can

141
00:08:33,920 --> 00:08:37,120
Speaker 1: based team had it covered. By the time they were finished,

142
00:08:37,160 --> 00:08:40,280
Speaker 1: the Bakery Scan platform was able to recognize dozens of

143
00:08:40,360 --> 00:08:43,960
Speaker 1: varieties of pastries with ninety eight percent accuracy, and it

144
00:08:44,040 --> 00:08:46,760
Speaker 1: was always learning too. The more the cashiers helped in

145
00:08:46,840 --> 00:08:50,320
Speaker 1: identifying questionable scans at the checkout line, the better the

146
00:08:50,320 --> 00:08:54,360
Speaker 1: program got it recognizing them the next time around. Bakery

147
00:08:54,400 --> 00:08:57,200
Speaker 1: Scan was a breakthrough, but it was only the beginning.

148
00:08:57,640 --> 00:09:00,319
Speaker 1: Another challenge was on the horizon, and it was about

149
00:09:00,320 --> 00:09:03,280
Speaker 1: to take Cambay's brain from the conveyor belt to the

150
00:09:03,320 --> 00:09:07,640
Speaker 1: oncology ward you see. Ten years after bakery Scans inception,

151
00:09:07,760 --> 00:09:11,240
Speaker 1: a doctor at Kyoto's Louis Pasture Center for Medical Research

152
00:09:11,559 --> 00:09:14,200
Speaker 1: saw a story about it on the news. The concept

153
00:09:14,240 --> 00:09:17,400
Speaker 1: intrigued him as someone who spent years looking at cancer

154
00:09:17,480 --> 00:09:21,080
Speaker 1: cells up close, because under the microscope they kind of

155
00:09:21,080 --> 00:09:24,720
Speaker 1: looked like bread, and he wondered if Canby's technology could

156
00:09:24,720 --> 00:09:29,480
Speaker 1: be used outside of the bakery and inside his research facility.

157
00:09:29,920 --> 00:09:33,560
Speaker 1: So Brain modified the algorithms, and with the little engineering,

158
00:09:33,760 --> 00:09:37,640
Speaker 1: the new platform dubbed AI scan could now detect a

159
00:09:37,720 --> 00:09:40,880
Speaker 1: single cancer cell among a microscope slide full of other

160
00:09:41,000 --> 00:09:46,280
Speaker 1: healthier cells with accuracy. Today, this AI tech is being

161
00:09:46,400 --> 00:09:49,920
Speaker 1: used across all sorts of different industries like automotive and

162
00:09:49,960 --> 00:09:53,440
Speaker 1: even jet engineering, and it's still saving lives in the

163
00:09:53,480 --> 00:09:57,880
Speaker 1: medical field too, all thanks to a complicated request from

164
00:09:57,880 --> 00:10:00,760
Speaker 1: a bakery. And I think the big your lesson here

165
00:10:00,960 --> 00:10:05,040
Speaker 1: speaks for itself. Don't ever let people tell you that

166
00:10:05,160 --> 00:10:10,679
Speaker 1: pastries aren't good for you. I hope you've enjoyed today's

167
00:10:10,679 --> 00:10:14,320
Speaker 1: guided tour of the Cabinet of Curiosities. Subscribe for free

168
00:10:14,400 --> 00:10:17,040
Speaker 1: on Apple Podcasts, or learn more about the show by

169
00:10:17,160 --> 00:10:22,120
Speaker 1: visiting Curiosities podcast dot com. The show was created by

170
00:10:22,160 --> 00:10:25,800
Speaker 1: me Aaron Manky in partnership with how Stuff Works. I

171
00:10:25,880 --> 00:10:29,680
Speaker 1: make another award winning show called Lore, which is a podcast,

172
00:10:29,800 --> 00:10:32,720
Speaker 1: book series, and television show, and you can learn all

173
00:10:32,760 --> 00:10:36,320
Speaker 1: about it over at the World of Lore dot com.

174
00:10:36,320 --> 00:10:39,640
Speaker 1: And until next time, stay curious. Ye