What should we ask AI to build? | Luqman Deen | TEDxGC Negombo Youth — Transcript

Full transcript

1. 0:00[music]
2. 0:07[applause]
3. 0:15>> I have a confession.
4. 0:17I used AI when drafting my initial ideas
5. 0:20for this speech.
6. 0:22But what it did not do is decide what I
7. 0:25want to say.
8. 0:26That decision was still mine.
9. 0:30This small experience made me think
10. 0:31about a much bigger picture.
11. 0:34As machines become more capable, what
12. 0:36responsibilities should remain with us?
13. 0:40AI is developing quickly.
14. 0:43Its growth is inevitable.
15. 0:45The use of it is usually frowned upon,
16. 0:48but it should be considered a privilege.
17. 0:51It may influence our cities,
18. 0:53education,
19. 0:55health care,
20. 0:56and the way we respond to environmental
21. 0:58challenges.
22. 1:00So, instead of asking only, what can AI
23. 1:02do?
24. 1:03Perhaps we should ask, what should we do
25. 1:06with it?
26. 1:08My first question is,
27. 1:10what can AI
28. 1:12AI could develop the smartest cities.
29. 1:15It could help plan bus and train routes,
30. 1:17reduce traffic, travel time, and fuel
31. 1:20use.
32. 1:22It could support doctors by helping them
33. 1:24analyze information and visualize their
34. 1:26problems more clearly.
35. 1:28It could help communities prepare for
36. 1:29disasters by identifying areas that may
37. 1:32need support.
38. 1:35But there is an important problem.
39. 1:37Making something faster, cheaper, or
40. 1:40more efficient does not automatically
41. 1:42make society better.
42. 1:45A city could have excellent technology,
43. 1:48but still be unfair.
44. 1:50An education system could use AI, but
45. 1:53still leave students behind.
46. 1:57A decision could look good on paper
47. 1:59while ignoring the people affected by
48. 2:01it.
49. 2:02For example,
50. 2:04if a school system used
51. 2:06AI
52. 2:08to decide which students needed extra
53. 2:09support,
54. 2:11the system might miss someone if its
55. 2:13information is incomplete.
56. 2:15The technology may not intend to be
57. 2:17unfair,
58. 2:19but the
59. 2:20result could still be unfair.
60. 2:22This is why we cannot judge a solution
61. 2:25only by how quickly it works.
62. 2:28We must also ask whether it works well
63. 2:31for the people.
64. 2:34My second question is, what should we
65. 2:36ask AI to build?
66. 2:38Instead of only asking AI to make things
67. 2:41faster, cheaper, and efficient,
68. 2:44we should ask how it can help create a
69. 2:46more educated society,
70. 2:49a healthier society,
71. 2:50a safer society,
72. 2:52and a more sustainable society.
73. 2:56The difference is that we start with a
74. 2:58human need instead of the technology
75. 3:00itself.
76. 3:01First,
77. 3:02we identify a real problem.
78. 3:05Second, we understand who is affected by
79. 3:08it.
80. 3:09And after that, we decide whether AI is
81. 3:12the right tool.
82. 3:15Sometimes it will be, sometimes not.
83. 3:18Sometimes a simpler solution, such as
84. 3:20better planning or listening to the
85. 3:22community, may work better.
86. 3:25We should not use AI just because it's
87. 3:28new or impressive.
88. 3:30We should also ask who might be left
89. 3:33out.
90. 3:34If a tool only works well for people
91. 3:37with fast internet, modern devices, or
92. 3:40particular language skills,
93. 3:42it may not benefit everyone equally.
94. 3:46A technology can be advanced and still
95. 3:49be inaccessible.
96. 3:52Progress should not only mean that some
97. 3:54people gain new advantages.
98. 3:57It should mean
99. 3:58that everyone
100. 4:00has a fair chance to benefit.
101. 4:05My final question is,
102. 4:07who decides what a better society means?
103. 4:11AI does not decide what makes a good
104. 4:13society.
105. 4:15People set its goals, choose how it's
106. 4:17used, and decide which outcomes really
107. 4:20matter.
108. 4:22That means people must take the
109. 4:25responsibility
110. 4:26for the choices.
111. 4:28Sometimes our goals conflict.
112. 4:30Do we prioritize economic growth or
113. 4:33sustainability?
114. 4:35Freedom
115. 4:36or safety?
116. 4:38Convenience or privacy?
117. 4:41A city might produce energy very
118. 4:43efficiently,
119. 4:44but does that justify damaging the
120. 4:46environment?
121. 4:48A company might collect information to
122. 4:50personalize the service,
123. 4:52but how much privacy are people willing
124. 4:54to give up for this convenience?
125. 4:57These questions do not always have
126. 4:59simple answers.
127. 5:02This is where ethics become important.
128. 5:06We need to distinguish between what we
129. 5:08can achieve
130. 5:09and what we should achieve.
131. 5:12AI can help us reach that goal, but it
132. 5:14cannot decide whether that goal is fair
133. 5:17or worth pursuing.
134. 5:20We must ask, who benefits? Who could be
135. 5:23harmed?
136. 5:24And who should be responsible if
137. 5:26something goes wrong?
138. 5:29We also need to remember that AI can
139. 5:31make mistakes or provide misleading
140. 5:33information.
141. 5:35ChatGPT itself gives us a disclaimer.
142. 5:38We should not accept every answer simply
143. 5:41because a machine gives it to us.
144. 5:44We need to check information, question
145. 5:46results, and understand the limits to
146. 5:49the tools we use.
147. 5:52Using AI responsibly does not mean
148. 5:55rejecting it completely.
149. 5:57It means using it with care.
150. 6:01Responsibility is shared.
151. 6:04Governments can set fair rules,
152. 6:06companies can test their systems and be
153. 6:08honest about their limitations.
154. 6:10Schools can teach students to use AI
155. 6:12thoughtfully.
156. 6:14And each of us should be able to
157. 6:16question the information we we receive
158. 6:19instead of accepting it without
159. 6:21thinking.
160. 6:24Ultimately,
161. 6:26AI should not decide the future for us.
162. 6:28It should be a tool that helps us build
163. 6:31the future we choose.
164. 6:34It can offer ideas, analyze information,
165. 6:37and help us find possible solutions.
166. 6:39But it cannot replace our responsibility
167. 6:43to decide what is fair,
168. 6:45what is important, and what kind of
169. 6:47society we want to be create.
170. 6:51AI may give us more and more ways to
171. 6:52solve problems.
172. 6:54But a better society is not simply one
173. 6:57with the most advanced technology.
174. 6:59It is one that uses the technology
175. 7:01thoughtfully, treats people fairly, and
176. 7:04considers the future.
177. 7:07The important question is not just
178. 7:09whether AI can build something.
179. 7:11It is whether what we build will
180. 7:13genuinely make life better for the
181. 7:16people. Thank you.