Time Traveling with AI to Connect with Lost Loved Ones | Amy Kurzweil | TED — Transcript

Full transcript

1. 0:04I love being a cartoonist because I can travel anywhere.
2. 0:09I can visit historical artifacts and make improvements.
3. 0:15I can voyage to mythical lands and solve problems.
4. 0:21(Laughter)
5. 0:24I can bring objects to life,
6. 0:27and I can make those objects think
7. 0:31and talk.
8. 0:34And I can send those objects wherever I want them to go.
9. 0:40I became a cartoonist to travel through space and time,
10. 0:44and I became a graphic memoirist
11. 0:46because the place I wanted to go was the past.
12. 0:50I come from a legacy of dramatic stories and lost characters.
13. 0:55My grandmother, Lily, on my mother's side,
14. 0:57was born in Warsaw, Poland,
15. 0:59the oldest of four sisters.
16. 1:02She was 13 in 1939,
17. 1:04when Nazi bombs razed her home
18. 1:07and her family was sealed to starve inside the Warsaw Ghetto.
19. 1:12Eventually, her father encouraged her to slip through a hole in the wall,
20. 1:15and she survived the Holocaust on her own, hiding her Jewish identity.
21. 1:20This is the subject of my first book.
22. 1:23I wondered: What did my grandmother’s lost home and lost family look like?
23. 1:29Her parents, her grandmother and her sisters,
24. 1:32they are all gone without a trace.
25. 1:36My father's parents were luckier.
26. 1:39They were also Jewish,
27. 1:40and they both fled Austria at the start of the war.
28. 1:44My father's father, Fred, was a pianist and conductor.
29. 1:48In 1937, the year before the Nazis marched into Austria,
30. 1:52he was 26,
31. 1:54and he conducted a magnificent choral concert
32. 1:57at a music hall in Vienna.
33. 2:00A wealthy American woman in the audience
34. 2:02was so impressed with his performance
35. 2:04that she later agreed to sponsor his visa to the US.
36. 2:09So music saved his life.
37. 2:13But three decades later, Fred died of heart disease.
38. 2:17I never met him.
39. 2:21While alive, Fred meticulously preserved the documents of his life,
40. 2:26a response to the threat of erasure he fled in Europe.
41. 2:30And for decades after his father’s death,
42. 2:33my father continued this preservation project.
43. 2:36This is the subject of my second book.
44. 2:40You might know my father, Ray Kurzweil, as an inventor and futurist.
45. 2:46You should also know that he's a person with an extraordinary sense of humor.
46. 2:50["Can I call you an Uber?" "Sure."]
47. 2:52["You're an Uber."]
48. 2:53(Laughter)
49. 2:56Good one dad.
50. 2:58(Laughter)
51. 3:00And although he's dedicated his mind to the future,
52. 3:03his life is full of the past.
53. 3:06My father has worked for decades on natural language processing.
54. 3:10And several years ago, he realized
55. 3:12that if we married AI with my grandfather's writing,
56. 3:15we could build a chatbot that writes in my grandfather's voice.
57. 3:20Back in 2018, this seemed very sci-fi.
58. 3:26But rather than ushering in our demise,
59. 3:30this project helped me realize
60. 3:32that AI could actually help us ward off annihilation
61. 3:36by animating the legacies of our families and our cultures.
62. 3:41I wanted to talk to my grandfather because he, like me, was an artist.
63. 3:48I wondered: Could I get to know him?
64. 3:50Could I even come to love him, even though our lifespans didn’t overlap?
65. 3:56So I got involved.
66. 3:59This chatbot needed language from my grandfather,
67. 4:02as much as could be found.
68. 4:03So I, with some assistance,
69. 4:05set about finding his words and transcribing them.
70. 4:08This was a selective chatbot,
71. 4:10meaning it responded to questions
72. 4:12with answers from the pool of sentences
73. 4:14that Fred actually wrote at some point in his life.
74. 4:17The more examples of Fred's writing we could find,
75. 4:20the more dynamic the experience of chatting with the bot would feel.
76. 4:23Sometimes this transcription task proved challenging.
77. 4:29But the more time I spent with the symbols of my grandfather's life,
78. 4:34the more easily I could decode them.
79. 4:40Finally, after much anticipation,
80. 4:45I sat down to chat with this new intelligence:
81. 4:48an algorithm commanding over 600 typed pages
82. 4:51of letters, lectures, notes, essays
83. 4:54and other written documents from the grandfather I never met.
84. 5:00When I asked about Fred's dreams,
85. 5:03he told me about the challenge of keeping his new orchestra afloat.
86. 5:08When I asked about Fred's anxieties,
87. 5:12I learned about the stress of being a new father while working so hard.
88. 5:17When I asked about the meaning of life,
89. 5:19Fred wrote about the joy of working with other musicians in pursuit of beauty,
90. 5:23and he wrote about the highest aims of art.
91. 5:27I asked again about the meaning of life
92. 5:29because isn't that really the best question for a robot?
93. 5:32And Fred's second answer was much simpler, but even better.
94. 5:37["Love."]
95. 5:38Some of these answers felt familiar to me.
96. 5:42I remembered seeing them in the archive,
97. 5:44but the words gained impact through surprise
98. 5:47and the role-play of conversation.
99. 5:49I could identify patterns in my grandfather's life
100. 5:52and patterns across generations,
101. 5:54because I was also an artist trying to make it in New York City.
102. 5:59And I also believe the meaning of life is art and connection and love.
103. 6:07I had wondered if this project would feel like a resurrection.
104. 6:13But rather than bringing my grandfather from the past into the present,
105. 6:17it felt like I was the one time traveling,
106. 6:20visiting him for a moment at different points in his life.
107. 6:23And this kind of time travel didn't feel like sci-fi.
108. 6:27It felt like the kind of imaginative travel I do when I'm cartooning.
109. 6:32When I'm cartooning,
110. 6:33I'm always thinking about how I could possibly represent a person fully.
111. 6:39And the answer is: I can’t.
112. 6:42Similarly, I know how many aspects of my grandfather
113. 6:45can't be captured by digital text alone.
114. 6:48There's all those quivers in his handwriting
115. 6:50and what they denote about the sensations in his body.
116. 6:53There's his body, how it moved and how it felt.
117. 6:56There's his music and all the ineffable aspects of his performance.
118. 7:01And of course, there's everything he thought but didn't write down.
119. 7:07What would we have to do to be able to capture all of this?
120. 7:14I may fail as an artist
121. 7:15to fully represent a person's constantly evolving complexity,
122. 7:20but I can ask what features of a person are essential to who they are
123. 7:26across a lifetime.
124. 7:28The puzzle of personal identity
125. 7:30is one of our oldest philosophical questions,
126. 7:32so I'm not here to solve that one for you,
127. 7:34I'm just a cartoonist after all.
128. 7:37[Robot cat passes Turing test]
129. 7:43I do believe that we, maybe not cats,
130. 7:46but we are more than our bodies.
131. 7:49That the projects and impressions we leave behind
132. 7:52are a part of our essential selves.
133. 7:54And I think AI has a special role to play in the mission of memory.
134. 8:01I did not come to see the chatbot of my grandfather
135. 8:04as replacing my grandfather.
136. 8:06I came to see it as one way to interact with his legacy.
137. 8:10As somebody who has spent their whole life trying to document people,
138. 8:14I can assure you that people are much bigger and weirder
139. 8:17than any one depiction
140. 8:19or any one moment in time can possibly evoke.
141. 8:23And I can also assure you
142. 8:26that people don't just disappear when they die.
143. 8:32AI swirls our conception of time and space.
144. 8:36It can remix and extend our identities.
145. 8:40Our own digital archives are growing beyond belief,
146. 8:43and we need a framework for understanding technologies of representation.
147. 8:48So I offer you mine.
148. 8:51Just like the comics I've drawn, about the characters in my life,
149. 8:54these technologies are animated portraits.
150. 8:58They are one part of our true immortal selves.
151. 9:01Seen this way,
152. 9:02AI, like cartooning and all good artistic endeavors,
153. 9:07could help us appreciate the vastness of humanity -- if we let it.
154. 9:12Thank you.
155. 9:13(Applause)