1
00:00:00,280 --> 00:00:04,920
Speaker 1: Is our destiny coded in the algorithm. If you feel

2
00:00:04,960 --> 00:00:08,480
Speaker 1: addicted to social media, this video is for you. If

3
00:00:08,520 --> 00:00:11,920
Speaker 1: you feel glued to whatever's on your feed and can't

4
00:00:11,960 --> 00:00:15,960
Speaker 1: stop doom scrolling, this video is for you. And if

5
00:00:15,960 --> 00:00:19,520
Speaker 1: you're worried about how social media is rewiring your brain,

6
00:00:20,040 --> 00:00:23,400
Speaker 1: this video is for you. Don't skip it. The number

7
00:00:23,400 --> 00:00:29,160
Speaker 1: one health and wellness podcast Jay Sety Jay Shetty set

8
00:00:31,560 --> 00:00:36,440
Speaker 1: I wanted to start today saying one thing. The algorithm

9
00:00:36,720 --> 00:00:39,879
Speaker 1: isn't as smart as we think it is. But the

10
00:00:39,960 --> 00:00:43,200
Speaker 1: deeper I went into my research, the more I realize

11
00:00:43,240 --> 00:00:48,520
Speaker 1: something unsettling. It's stronger than me, stronger than you, stronger

12
00:00:48,560 --> 00:00:52,960
Speaker 1: than all of us because it knows our weaknesses. But

13
00:00:53,080 --> 00:00:56,680
Speaker 1: here's what I also found. Even the strongest system that

14
00:00:56,680 --> 00:01:00,360
Speaker 1: has a glitch, the algorithm doesn't just know us. It

15
00:01:00,400 --> 00:01:03,639
Speaker 1: depends on us. And if we learn how it feeds,

16
00:01:04,120 --> 00:01:07,720
Speaker 1: we can decide whether to starve it or steer it.

17
00:01:08,440 --> 00:01:11,479
Speaker 1: When you google the words will I ever, the first

18
00:01:11,520 --> 00:01:15,000
Speaker 1: thing that comes up is will I ever find love?

19
00:01:15,840 --> 00:01:20,120
Speaker 1: The second is will I ever be enough? And the

20
00:01:20,160 --> 00:01:23,720
Speaker 1: third is will I am net worth? We go from

21
00:01:23,800 --> 00:01:28,000
Speaker 1: love to worth to money really quickly. But this search

22
00:01:28,040 --> 00:01:32,759
Speaker 1: for love, worth and belonging is what the algorithm exploits,

23
00:01:33,280 --> 00:01:37,760
Speaker 1: but not in the way you think. Picture this, it's midnight.

24
00:01:38,840 --> 00:01:42,480
Speaker 1: Think of a girl named Amelia lies in bed, phone

25
00:01:42,520 --> 00:01:46,440
Speaker 1: in her hand. She posts a photo, nothing dramatic, just

26
00:01:46,520 --> 00:01:50,960
Speaker 1: hoping someone notices. The likes trickle in her friend's comment.

27
00:01:51,400 --> 00:01:56,120
Speaker 1: She taps on another girl's profile, prettier, thinner, more followers,

28
00:01:56,840 --> 00:02:01,080
Speaker 1: She lingers, she clicks, She scrolls of them, pays attention.

29
00:02:02,040 --> 00:02:06,360
Speaker 1: The next night, her feed feels different, more flawless faces,

30
00:02:06,720 --> 00:02:11,760
Speaker 1: more filters, more diets, more lives that look nothing like hers.

31
00:02:12,360 --> 00:02:18,520
Speaker 1: Curiosity turns into comparison. Comparison turns into obsession, and soon

32
00:02:18,800 --> 00:02:23,720
Speaker 1: every scroll feels like it's whispering the same three words,

33
00:02:24,760 --> 00:02:30,799
Speaker 1: You're not enough. Until one night she doesn't see herself anymore.

34
00:02:31,560 --> 00:02:35,680
Speaker 1: She only sees the mirror the algorithm is holding up

35
00:02:35,720 --> 00:02:40,000
Speaker 1: to her. This isn't just a Merelia story. Fifty six

36
00:02:40,040 --> 00:02:42,800
Speaker 1: percent of girls feel they can't live up to the

37
00:02:42,840 --> 00:02:46,960
Speaker 1: beauty standards they see on social media. Ninety percent of

38
00:02:47,000 --> 00:02:50,920
Speaker 1: girls follow at least one social media account that makes

39
00:02:50,960 --> 00:02:55,440
Speaker 1: them feel less beautiful. But here's the real question. Did

40
00:02:55,440 --> 00:02:59,440
Speaker 1: the algorithm build that mirror or did she was it

41
00:02:59,520 --> 00:03:03,720
Speaker 1: coded in Silicon Valley or coded in our own clicks.

42
00:03:04,160 --> 00:03:08,720
Speaker 1: Let's look at the algorithm first. What do algorithms actually do?

43
00:03:09,400 --> 00:03:12,560
Speaker 1: Number one They watch every pause, every click, every like,

44
00:03:12,639 --> 00:03:16,320
Speaker 1: every share, even how long you hover over a video

45
00:03:16,600 --> 00:03:21,000
Speaker 1: or comment. TikTok tracks watch time down to the second.

46
00:03:21,480 --> 00:03:24,640
Speaker 1: If you rewatch a clip, it's a super strong signal.

47
00:03:25,320 --> 00:03:29,120
Speaker 1: Number two. They predict using your history and the behaviors

48
00:03:29,160 --> 00:03:33,240
Speaker 1: of millions of people like you, Algorithms predict what are

49
00:03:33,320 --> 00:03:36,800
Speaker 1: you most likely to engage with. Next. If people who

50
00:03:36,840 --> 00:03:40,640
Speaker 1: watch fitness videos also tend to watch diet hacks, you'll

51
00:03:40,680 --> 00:03:45,800
Speaker 1: probably get diet hacks. Number three. They amplify. The posts

52
00:03:45,800 --> 00:03:50,600
Speaker 1: that get more engagement, especially emotional engagement, are pushed to

53
00:03:50,800 --> 00:03:56,200
Speaker 1: more people. Number four. They adapt. Every click retrains the system.

54
00:03:56,680 --> 00:03:59,920
Speaker 1: Your feed tomorrow is shaped by what you do today.

55
00:04:00,520 --> 00:04:06,240
Speaker 1: YouTube's recommendation engine is called a reinforcement system. It's literally

56
00:04:06,280 --> 00:04:10,200
Speaker 1: designed to learn from your actions in real time. The

57
00:04:10,240 --> 00:04:13,920
Speaker 1: most accurate model is a cycle. First of all, we

58
00:04:14,040 --> 00:04:17,839
Speaker 1: click what feels good, familiar, or emotionally hot. Two, the

59
00:04:17,960 --> 00:04:20,960
Speaker 1: algorithm learns and serves us more of that to keep

60
00:04:21,040 --> 00:04:24,919
Speaker 1: us there. Number three, we become more entrenched and less

61
00:04:24,920 --> 00:04:29,799
Speaker 1: exposed to alternatives, and number four, outraging divisions spread faster

62
00:04:30,200 --> 00:04:35,680
Speaker 1: because anger is more contagious. In plain words, the algorithm

63
00:04:35,800 --> 00:04:39,799
Speaker 1: isn't a mastermind. It's a machine that asks one question,

64
00:04:40,279 --> 00:04:45,720
Speaker 1: over and over again, what will keep you here the longest?

65
00:04:46,680 --> 00:04:50,640
Speaker 1: It's like a maximum security prison. So how do we

66
00:04:50,680 --> 00:04:56,520
Speaker 1: get trapped? First? The nudge think, Netflix, AutoPlay, TikTok, infinite scroll,

67
00:04:56,920 --> 00:05:00,880
Speaker 1: the design that says, don't think, don't choose, just keep watching.

68
00:05:01,440 --> 00:05:04,520
Speaker 1: That's how you start a Korean baking show you didn't

69
00:05:04,520 --> 00:05:08,159
Speaker 1: even know existed, and three hours later you're crying over

70
00:05:08,200 --> 00:05:13,560
Speaker 1: a documentary on penguins. A study found disabling AutoPlay led

71
00:05:13,600 --> 00:05:18,760
Speaker 1: to a seventeen minute shorter average session. Showing AutoPlay measurably

72
00:05:18,800 --> 00:05:23,479
Speaker 1: extends watch time. It's not a choice disappearing. It's a

73
00:05:23,600 --> 00:05:29,520
Speaker 1: choice so well hidden you don't realize you never made it. Second,

74
00:05:29,600 --> 00:05:34,560
Speaker 1: the loop, Yale researchers found when people post moral outrage online,

75
00:05:35,000 --> 00:05:39,160
Speaker 1: people reward them with likes and retweets. That person now

76
00:05:39,200 --> 00:05:42,960
Speaker 1: posts even more outrage the next time. It's not the algorithm,

77
00:05:42,960 --> 00:05:46,280
Speaker 1: it's us. It's real people. As one researcher put it,

78
00:05:46,640 --> 00:05:50,919
Speaker 1: we don't just consume outrage, we start producing it because

79
00:05:51,080 --> 00:05:58,039
Speaker 1: outrage performs better than honesty. And Third, the push Mozilla's

80
00:05:58,080 --> 00:06:03,160
Speaker 1: YouTube Regrets project from twenty found that volunteers who started

81
00:06:03,200 --> 00:06:08,720
Speaker 1: with neutral searches like fitness or politics reported being steered

82
00:06:09,000 --> 00:06:15,000
Speaker 1: toward extremist, conspiratorial, or misogynistic content. Seventy one percent of

83
00:06:15,040 --> 00:06:19,800
Speaker 1: the videos people regretted watching were never searched for they

84
00:06:19,800 --> 00:06:24,040
Speaker 1: were recommended the ucl Kent study from twenty twenty four.

85
00:06:24,520 --> 00:06:28,760
Speaker 1: In a recent algorithmic model study, accounts on TikTok were

86
00:06:28,800 --> 00:06:32,560
Speaker 1: shown four times more misogynistic content on the for you

87
00:06:32,720 --> 00:06:37,279
Speaker 1: page within just five days of casual strolling. What does

88
00:06:37,320 --> 00:06:42,000
Speaker 1: this do to men and women? Women get more insecure

89
00:06:42,040 --> 00:06:47,320
Speaker 1: about their appearance. Men get more exposed to misogynistic content.

90
00:06:47,960 --> 00:06:52,120
Speaker 1: Women experience more anxiety and self doubt. Men become more

91
00:06:52,160 --> 00:06:56,440
Speaker 1: lonely and disconnected. Women compared their lives to others and

92
00:06:56,480 --> 00:07:00,840
Speaker 1: feel their falling behind. Men compared their state to others

93
00:07:01,360 --> 00:07:04,760
Speaker 1: and feel like they're being left behind. Both end up

94
00:07:04,760 --> 00:07:09,520
Speaker 1: in the same place on social media, isolated, exhausted, and

95
00:07:09,600 --> 00:07:14,000
Speaker 1: shaped by the same machine. The algorithm will do anything

96
00:07:14,080 --> 00:07:18,280
Speaker 1: to keep us glued. There is a huge incentive issue

97
00:07:18,280 --> 00:07:21,120
Speaker 1: for the algorithm because in one study where they chose

98
00:07:21,160 --> 00:07:26,080
Speaker 1: not to show toxic posts, users spent approximately nine percent

99
00:07:26,200 --> 00:07:31,360
Speaker 1: less time daily, experience fewer AD impressions, and generated fewer

100
00:07:31,440 --> 00:07:36,400
Speaker 1: AD clicks. The algorithm's goal is not to make us polarized.

101
00:07:36,960 --> 00:07:39,760
Speaker 1: It's not to make us happy. It's to make us

102
00:07:39,760 --> 00:07:43,560
Speaker 1: addicted and glued to our screens. It is showing you

103
00:07:43,680 --> 00:07:47,680
Speaker 1: what people like you are engaging with, assuming you will

104
00:07:47,760 --> 00:07:51,320
Speaker 1: stay as well. We talked about what the algorithm does.

105
00:07:51,960 --> 00:07:55,560
Speaker 1: Let's look at what role we play. Our clicks build

106
00:07:55,560 --> 00:07:59,960
Speaker 1: the cage. False news stories are seventy percent more likely

107
00:08:00,160 --> 00:08:04,640
Speaker 1: to be retweeted than true stories are. It also takes

108
00:08:04,680 --> 00:08:08,239
Speaker 1: true stories about six times as long to reach fifteen

109
00:08:08,360 --> 00:08:12,040
Speaker 1: hundred people as it does for false stories to reach

110
00:08:12,120 --> 00:08:16,160
Speaker 1: the same number of people. Algorithms don't see truth or lies,

111
00:08:16,520 --> 00:08:28,240
Speaker 1: They only see clicks from people like us want to

112
00:08:28,280 --> 00:08:31,800
Speaker 1: make a real difference This giving season, this December on

113
00:08:31,920 --> 00:08:36,400
Speaker 1: Purpose is part of Pods Fight Poverty podcast, teaming up

114
00:08:36,440 --> 00:08:40,320
Speaker 1: to lift three villages in Rwanda out of extreme poverty.

115
00:08:40,720 --> 00:08:44,280
Speaker 1: We're doing it through Give Directly, which sends cash straight

116
00:08:44,320 --> 00:08:47,439
Speaker 1: to families so they can choose what they need most.

117
00:08:48,000 --> 00:08:52,839
Speaker 1: Donate at GiveDirectly dot org forward slash on purpose. First

118
00:08:52,880 --> 00:08:56,320
Speaker 1: time gifts are matched, doubling your impact. Our goal is

119
00:08:56,440 --> 00:09:00,720
Speaker 1: one million dollars by year's end, enough to lift seven

120
00:09:00,800 --> 00:09:05,400
Speaker 1: hundred families out of poverty. Join us at GiveDirectly dot

121
00:09:05,520 --> 00:09:18,160
Speaker 1: org forward slash on purpose. Number two. False news spread

122
00:09:18,280 --> 00:09:22,720
Speaker 1: six times faster than true news because shocking content sparks

123
00:09:22,760 --> 00:09:26,160
Speaker 1: more clicks and shares from us, so the algorithm promotes

124
00:09:26,200 --> 00:09:31,120
Speaker 1: it further. The content must already have emotional potency. An

125
00:09:31,120 --> 00:09:35,720
Speaker 1: algorithm won't manufacture depth or resonance from nothing. It can't

126
00:09:35,880 --> 00:09:39,800
Speaker 1: make it go viral. Number three. For major media outlets,

127
00:09:40,160 --> 00:09:44,800
Speaker 1: each additional negative effect word in a post is associated

128
00:09:44,960 --> 00:09:48,160
Speaker 1: with a five to eight percent increase in shares and

129
00:09:48,200 --> 00:09:53,240
Speaker 1: retweets from US and four Facebook studies showed that even

130
00:09:53,280 --> 00:09:58,480
Speaker 1: when given the opportunity, users click links confirming their bias

131
00:09:59,000 --> 00:10:03,800
Speaker 1: far more often than opposing one. Liberals chose cross cutting

132
00:10:03,840 --> 00:10:07,880
Speaker 1: news twenty one percent of the time, conservatives thirty percent

133
00:10:07,920 --> 00:10:11,840
Speaker 1: of the time. Here's the twist. The algorithm doesn't pick

134
00:10:11,920 --> 00:10:16,160
Speaker 1: sides we do. It just learns our choice and builds

135
00:10:16,160 --> 00:10:19,320
Speaker 1: a fortress around it. The danger is in that we

136
00:10:19,400 --> 00:10:22,640
Speaker 1: have no choice. It's that we don't notice when our

137
00:10:22,720 --> 00:10:26,599
Speaker 1: choices are being shaped for us. So let's do a

138
00:10:26,640 --> 00:10:30,880
Speaker 1: thought experiment. Why don't we create a social media platform

139
00:10:31,200 --> 00:10:35,120
Speaker 1: without these incentives, one that doesn't play these games with us.

140
00:10:36,200 --> 00:10:39,079
Speaker 1: They already tried that, and when I'm about to share

141
00:10:39,120 --> 00:10:42,920
Speaker 1: with you shocked me the most. A new study out

142
00:10:42,960 --> 00:10:46,440
Speaker 1: of the University of Amsterdam tested this by creating a

143
00:10:46,480 --> 00:10:51,720
Speaker 1: stripped down social network, no ads, no recommendation algorithms, no

144
00:10:51,880 --> 00:10:58,040
Speaker 1: invisible hand pushing content. Researchers released five hundred AI chatbots

145
00:10:58,120 --> 00:11:02,120
Speaker 1: onto the platform, each powered by Open Ai, and gave

146
00:11:02,160 --> 00:11:06,439
Speaker 1: them distinct political and social identities. Then they let them

147
00:11:06,440 --> 00:11:11,160
Speaker 1: loose across five separate experiments amounting to ten thousand interactions.

148
00:11:11,520 --> 00:11:16,319
Speaker 1: The bots began to behave exactly like us. They followed

149
00:11:16,360 --> 00:11:19,959
Speaker 1: those who thought like them. They've reposted the loudest, most

150
00:11:20,000 --> 00:11:24,640
Speaker 1: extreme voices. They gravitated into echo chambers, not because an

151
00:11:24,720 --> 00:11:28,160
Speaker 1: algorithm pushed them there, but because that's where they chose

152
00:11:28,200 --> 00:11:31,679
Speaker 1: to go. It also found that users who posted the

153
00:11:31,720 --> 00:11:36,199
Speaker 1: most partisan content tended to get the most followers and reposts.

154
00:11:37,040 --> 00:11:43,600
Speaker 1: Researchers tried interventions dampening virality, hiding follow accounts, even boosting

155
00:11:43,640 --> 00:11:48,560
Speaker 1: opposing views, but nothing cracked the cycle. The most they

156
00:11:48,640 --> 00:11:53,360
Speaker 1: managed was a six percent reduction in partisan engagement. In

157
00:11:53,400 --> 00:11:57,360
Speaker 1: some cases, when they started hiding user bios, the divide

158
00:11:57,400 --> 00:12:01,520
Speaker 1: actually grew sharper, and the most extreme posts gained even

159
00:12:01,559 --> 00:12:06,040
Speaker 1: more traction. The implication is chilling. Maybe it isn't just

160
00:12:06,080 --> 00:12:10,480
Speaker 1: the algorithms that wop us. Maybe social media itself is

161
00:12:10,559 --> 00:12:13,640
Speaker 1: wired against our better nature. Think about it like a

162
00:12:13,679 --> 00:12:17,520
Speaker 1: funhouse mirror. It doesn't simply reflect who we are. It

163
00:12:17,640 --> 00:12:21,400
Speaker 1: stretches our fears. It magnifies our biases and turns our

164
00:12:21,440 --> 00:12:26,520
Speaker 1: flaws into a spectacle. As humans, we can live consciously

165
00:12:27,320 --> 00:12:31,800
Speaker 1: or unconsciously. We can choose our stronger selves or our

166
00:12:31,880 --> 00:12:36,480
Speaker 1: weaker selves. When we choose our weaker self, humans are

167
00:12:36,480 --> 00:12:40,959
Speaker 1: not just curious. We're programmed to measure ourselves against others.

168
00:12:42,000 --> 00:12:47,120
Speaker 1: Comparison is our oldest survival instinct. Envy is the emotional fuel.

169
00:12:48,000 --> 00:12:51,199
Speaker 1: The algorithm didn't invent it, but it does exploit it

170
00:12:51,679 --> 00:12:56,360
Speaker 1: when we're tired, overwhelmed, and exhausted. Humans are not ruled

171
00:12:56,360 --> 00:13:01,040
Speaker 1: by curiosity. We're ruled by comparison, and envy is the

172
00:13:01,120 --> 00:13:05,640
Speaker 1: price of admission. The algorithm didn't create envy, it just

173
00:13:05,720 --> 00:13:09,240
Speaker 1: turned it into an economy. Now why do we do this?

174
00:13:10,000 --> 00:13:14,840
Speaker 1: The first is negativity bias. Evolution tuned us to notice

175
00:13:14,920 --> 00:13:19,760
Speaker 1: threats more than opportunities. Missing a bery was fine, Missing

176
00:13:19,800 --> 00:13:25,120
Speaker 1: a snake was fatal. Number two, outrage is social currency.

177
00:13:26,000 --> 00:13:30,480
Speaker 1: Expressing outrage signals loyalty to your group. It tells others

178
00:13:30,679 --> 00:13:34,920
Speaker 1: I'm one of us, and in polarized contexts, this isn't

179
00:13:35,000 --> 00:13:40,840
Speaker 1: just emotion, its identity signaling. Clicking rage is clicking belonging.

180
00:13:41,920 --> 00:13:46,960
Speaker 1: Number three cognitive efficiency. Negative content is often simpler. This

181
00:13:47,160 --> 00:13:51,640
Speaker 1: is bad. They're wrong, We're threatened. The brain prefers cognitive

182
00:13:51,720 --> 00:13:58,280
Speaker 1: ease over nuance. Complex balanced content demands more effort. Negativity

183
00:13:58,320 --> 00:14:02,800
Speaker 1: feels immedia, digest and actionable. So what do we do

184
00:14:02,920 --> 00:14:10,199
Speaker 1: about this? Doom Scrolling increases cortisol, anxiety, and learned helplessness.

185
00:14:10,240 --> 00:14:13,800
Speaker 1: In that state, people feel like they have no agency,

186
00:14:14,360 --> 00:14:17,760
Speaker 1: which can reinforce the sense of doom. So we have

187
00:14:17,800 --> 00:14:20,720
Speaker 1: an incentive issue for the platforms because they're just trying

188
00:14:20,720 --> 00:14:22,840
Speaker 1: to keep us glued, and we have a lack of

189
00:14:22,920 --> 00:14:27,920
Speaker 1: mental resilience for us. Put those both together, that's what

190
00:14:27,960 --> 00:14:30,960
Speaker 1: we're experiencing. Right now. So what do we do about

191
00:14:30,960 --> 00:14:35,000
Speaker 1: the incentive issue? People often ask me if I think

192
00:14:35,080 --> 00:14:38,760
Speaker 1: AI will ever have a soul, and my response is,

193
00:14:39,680 --> 00:14:42,160
Speaker 1: I don't know if AI will ever have a soul.

194
00:14:43,040 --> 00:14:46,600
Speaker 1: I just hope that people building AI have a soul.

195
00:14:47,480 --> 00:14:51,760
Speaker 1: The people who created these algorithms will lose millions or

196
00:14:51,880 --> 00:14:56,000
Speaker 1: billions if they adjusted the algorithm. Would they do that,

197
00:14:56,760 --> 00:15:00,720
Speaker 1: Will they recognize or think they have a responsibility. It's

198
00:15:00,720 --> 00:15:05,920
Speaker 1: a really interesting thing to think about because it's almost

199
00:15:05,960 --> 00:15:09,440
Speaker 1: like we're making something that is becoming us. It's almost

200
00:15:09,440 --> 00:15:14,680
Speaker 1: like Frankenstein, that idea that whatever system we build has

201
00:15:14,720 --> 00:15:17,520
Speaker 1: a part of us in it. If you build a company,

202
00:15:17,560 --> 00:15:19,240
Speaker 1: it has a part of you in it. There's an

203
00:15:19,480 --> 00:15:23,000
Speaker 1: energetic exchange as well. So what does that feel like

204
00:15:23,040 --> 00:15:27,040
Speaker 1: when you're building a platform that millions and billions of

205
00:15:27,080 --> 00:15:30,480
Speaker 1: people use. The truth is we can't afford to just

206
00:15:30,560 --> 00:15:34,640
Speaker 1: diagnose the problem. And I get intrigued by that sometimes

207
00:15:34,680 --> 00:15:36,760
Speaker 1: when people just want to diagnose the problem. But we

208
00:15:36,840 --> 00:15:41,320
Speaker 1: need to find solutions, and here are three changes social

209
00:15:41,360 --> 00:15:45,560
Speaker 1: media companies could try. The first is platforms should offer

210
00:15:45,680 --> 00:15:50,160
Speaker 1: chronological feeds by default, not buried in settings, and give

211
00:15:50,360 --> 00:15:56,720
Speaker 1: users transparent control to toggle between chronological and algorithmic. Facebook's

212
00:15:56,720 --> 00:16:03,600
Speaker 1: own studies show chronological feeds reduce political polarization and misinformation exposure,

213
00:16:04,000 --> 00:16:06,960
Speaker 1: though engagement does drop. The second thing they can do

214
00:16:07,160 --> 00:16:13,280
Speaker 1: is actually probably my favorite. Add friction before sharing. For example,

215
00:16:13,760 --> 00:16:18,240
Speaker 1: read before retweet prompts, share limits, cooling off periods on

216
00:16:18,320 --> 00:16:22,800
Speaker 1: viral posts. Imagine you couldn't share something until you add

217
00:16:22,840 --> 00:16:26,120
Speaker 1: read it in full. Imagine you can share something until

218
00:16:26,160 --> 00:16:47,120
Speaker 1: you'd what's that video in full? Twitter's twenty twenty read

219
00:16:47,200 --> 00:16:51,560
Speaker 1: before retweet experiment led to a forty percent increase in

220
00:16:51,640 --> 00:16:58,200
Speaker 1: people opening articles before sharing. WhatsApps forwarding limits dramatically slowed

221
00:16:58,320 --> 00:17:02,840
Speaker 1: misinformation in India. This could actually make a difference because

222
00:17:02,880 --> 00:17:08,040
Speaker 1: not only are we misinforming others, we're underinformed ourselves. If

223
00:17:08,080 --> 00:17:11,199
Speaker 1: you're retweeting something just based on the headline and have

224
00:17:11,280 --> 00:17:15,040
Speaker 1: no idea what's inside of it, we're now propelling ideas

225
00:17:15,440 --> 00:17:19,160
Speaker 1: that we don't fully grasp and understand. And number three

226
00:17:19,680 --> 00:17:25,440
Speaker 1: require algorithmic transparency and independent audits. Companies must publish how

227
00:17:25,480 --> 00:17:31,880
Speaker 1: recommendation systems prioritize content and allow external researchers to study

228
00:17:32,040 --> 00:17:36,919
Speaker 1: the impacts. The EU's Digital Services Act is already moving

229
00:17:36,960 --> 00:17:42,000
Speaker 1: this way, requiring large platforms to open their algorithms to scrutiny.

230
00:17:42,880 --> 00:17:46,000
Speaker 1: Now what do we do about the human nature issue?

231
00:17:46,520 --> 00:17:48,360
Speaker 1: I want to share with you one of my favorite stories.

232
00:17:49,400 --> 00:17:52,840
Speaker 1: A student once asked the Buddha, what do you gain

233
00:17:52,880 --> 00:17:57,760
Speaker 1: from meditation? He said nothing. The student asked them, why

234
00:17:57,840 --> 00:18:01,600
Speaker 1: do you meditate if you gain nothing? The Buddha replied,

235
00:18:02,119 --> 00:18:05,720
Speaker 1: I don't meditate because of what I gain. I meditate

236
00:18:06,200 --> 00:18:10,520
Speaker 1: because of what I lose. I lose anger, I lose envy,

237
00:18:10,920 --> 00:18:15,160
Speaker 1: I lose ego. If the algorithm is made of us,

238
00:18:15,960 --> 00:18:21,600
Speaker 1: then changing it doesn't start with code, It starts with character.

239
00:18:22,680 --> 00:18:26,359
Speaker 1: We have to remember that we are wired for generosity

240
00:18:27,040 --> 00:18:32,360
Speaker 1: but educated for greed. When will we finally start teaching

241
00:18:32,440 --> 00:18:37,159
Speaker 1: emotional mastery in schools? How long before we start teaching

242
00:18:37,200 --> 00:18:41,720
Speaker 1: critical thinking at an early age. Maybe the real test

243
00:18:42,359 --> 00:18:46,920
Speaker 1: isn't to build a happier network, it's to build happier users.

244
00:18:47,480 --> 00:18:50,639
Speaker 1: We built a machine to know us, and it became

245
00:18:50,760 --> 00:18:54,440
Speaker 1: us when we started on purpose. There are only three

246
00:18:54,480 --> 00:18:58,200
Speaker 1: things that went viral, cats and dogs. Sorry to put

247
00:18:58,240 --> 00:19:01,400
Speaker 1: them in the same group, babies and people taking their

248
00:19:01,400 --> 00:19:06,000
Speaker 1: clothes off. I had the innocent intention the naive vision

249
00:19:06,640 --> 00:19:10,879
Speaker 1: of making wisdom go viral. Today, we do over five

250
00:19:11,160 --> 00:19:16,280
Speaker 1: hundred million views across platforms every month, not playing into

251
00:19:16,359 --> 00:19:19,159
Speaker 1: rage bait, not trying to get pivot at be angry.

252
00:19:20,080 --> 00:19:22,800
Speaker 1: What does it show me? It shows me that people

253
00:19:22,840 --> 00:19:26,760
Speaker 1: will choose healthier options if they're available, if it's presented

254
00:19:26,800 --> 00:19:30,399
Speaker 1: to them in a digestible way. People will choose a

255
00:19:30,440 --> 00:19:33,159
Speaker 1: salad if they know why it's better for them, and

256
00:19:33,200 --> 00:19:36,920
Speaker 1: if it's available and has a great dressing. It's our

257
00:19:37,000 --> 00:19:40,040
Speaker 1: role to not play into the fear and find ways

258
00:19:40,080 --> 00:19:44,160
Speaker 1: to make love more attractive and accessible. It's so easy

259
00:19:44,200 --> 00:19:48,160
Speaker 1: to sell fear, It's so easy to sell negativity, it's

260
00:19:48,200 --> 00:19:52,040
Speaker 1: so easy to sell gossip. But the truth is, why

261
00:19:52,160 --> 00:19:56,960
Speaker 1: sell the things that sell people short? Why not provide

262
00:19:56,960 --> 00:20:01,800
Speaker 1: them with alternatives that are healthy, strengthening, empowering, that give

263
00:20:01,840 --> 00:20:03,919
Speaker 1: them the tools to make a difference in their life.

264
00:20:04,680 --> 00:20:09,280
Speaker 1: Here's the good news. Algorithms do not fully decide your fate.

265
00:20:09,960 --> 00:20:14,400
Speaker 1: They're predictive, not deterministic. They rely on your past clicks.

266
00:20:14,840 --> 00:20:19,160
Speaker 1: But you can override them by searching, subscribing to diverse sources,

267
00:20:19,560 --> 00:20:23,800
Speaker 1: and consciously engaging with content outside of your bubble. So

268
00:20:23,880 --> 00:20:25,479
Speaker 1: I want you to take a look at a new account.

269
00:20:25,520 --> 00:20:28,280
Speaker 1: I started in the for you page. The for you

270
00:20:28,400 --> 00:20:34,159
Speaker 1: page is pretty simple. It's beautiful imagery. It's introducing me

271
00:20:34,200 --> 00:20:37,720
Speaker 1: to a couple of scenery, and as I scroll down,

272
00:20:37,800 --> 00:20:40,400
Speaker 1: you start to see more of what the average person

273
00:20:40,440 --> 00:20:42,400
Speaker 1: would see. The for you page as you go deeper

274
00:20:42,960 --> 00:20:48,040
Speaker 1: shows you everything from political podcasts, shows me people working out,

275
00:20:48,080 --> 00:20:51,359
Speaker 1: shows me influencer content. Now I'm going to show you

276
00:20:51,400 --> 00:20:53,640
Speaker 1: how easy it is to change your for you page

277
00:20:54,200 --> 00:20:56,280
Speaker 1: because this page is so visual. I'm going to do

278
00:20:56,320 --> 00:20:58,760
Speaker 1: it through finding quotes and also, you know I love quotes,

279
00:21:00,119 --> 00:21:02,240
Speaker 1: So I'm going to go follow some quotes. I'm going

280
00:21:02,320 --> 00:21:06,000
Speaker 1: to like some quotes, going to like another quote, I'm

281
00:21:06,040 --> 00:21:07,879
Speaker 1: going to hover over it for a while. This is

282
00:21:07,920 --> 00:21:11,960
Speaker 1: really important to actually hover over the quote, to actually

283
00:21:12,040 --> 00:21:14,520
Speaker 1: read it, to actually be present with it. And now

284
00:21:14,520 --> 00:21:16,439
Speaker 1: I'm even going to share a quote with a friend

285
00:21:17,200 --> 00:21:18,840
Speaker 1: who's now going to think they have an issue because

286
00:21:18,840 --> 00:21:22,320
Speaker 1: I just shared some wisdom with them. When I refresh

287
00:21:22,960 --> 00:21:27,280
Speaker 1: check out my for you page, it's pretty much all quotes.

288
00:21:28,400 --> 00:21:32,399
Speaker 1: Through three to four simple steps, I transformed my for

289
00:21:32,560 --> 00:21:36,719
Speaker 1: you page. This is almost a cleansing, filtering process that

290
00:21:36,760 --> 00:21:39,960
Speaker 1: I recommend you do. It's simple. I want you to

291
00:21:40,000 --> 00:21:45,119
Speaker 1: follow five people you wouldn't usually follow. Agency isn't eliminated,

292
00:21:45,640 --> 00:21:49,880
Speaker 1: it's eroded by habit. People who intentionally curate their feeds,

293
00:21:50,160 --> 00:21:56,240
Speaker 1: limit usage, or diversify inputs show significantly less polarization. The

294
00:21:56,320 --> 00:21:58,720
Speaker 1: second thing I want you to do is hover over

295
00:21:59,200 --> 00:22:02,920
Speaker 1: and comment on five pieces of content you want to

296
00:22:02,960 --> 00:22:06,840
Speaker 1: see more of your offline life still matters. Real books,

297
00:22:06,880 --> 00:22:11,560
Speaker 1: real conversations, and communities can counteract the digital echo chamber.

298
00:22:12,359 --> 00:22:14,760
Speaker 1: And number three, I want you to share five pieces

299
00:22:14,800 --> 00:22:18,920
Speaker 1: of content you wouldn't usually and see how that changes

300
00:22:18,960 --> 00:22:22,480
Speaker 1: your algorithm. Number four, don't look at your phone first

301
00:22:22,480 --> 00:22:25,679
Speaker 1: thing in the morning. It's like letting one hundred strangers

302
00:22:25,720 --> 00:22:29,159
Speaker 1: walk into your bedroom before you've brush your teeth or

303
00:22:29,280 --> 00:22:32,240
Speaker 1: washed your face. You would never do that in real life,

304
00:22:32,800 --> 00:22:36,840
Speaker 1: don't do it online. And five be present with joy.

305
00:22:37,560 --> 00:22:42,440
Speaker 1: Celebrate your friend's wins and accomplishments. Stop overreacting to negativity

306
00:22:42,960 --> 00:22:47,720
Speaker 1: and underreacting to joy. We remember the bad times more

307
00:22:47,760 --> 00:22:51,280
Speaker 1: than the good times, because when we lose, we cry

308
00:22:51,359 --> 00:22:54,639
Speaker 1: for a month, and when we win, we celebrate for

309
00:22:54,680 --> 00:22:57,920
Speaker 1: a night. Here's what I want you to remember when

310
00:22:57,920 --> 00:23:02,240
Speaker 1: you like something. You're telling the algorithm show me more

311
00:23:02,280 --> 00:23:06,280
Speaker 1: of this. When you hover over something, you're saying to

312
00:23:06,359 --> 00:23:10,000
Speaker 1: the algorithm, I pay attention when you show me this.

313
00:23:10,600 --> 00:23:13,639
Speaker 1: When you comment on something, you're saying, this is really

314
00:23:13,680 --> 00:23:17,639
Speaker 1: important to me. And when you share it off the platform,

315
00:23:18,040 --> 00:23:23,640
Speaker 1: you're saying, fill my feed with this. You're co creating

316
00:23:23,920 --> 00:23:28,080
Speaker 1: your algorithm. You're actually coding it. One of my favorite

317
00:23:28,080 --> 00:23:32,840
Speaker 1: thoughts comes from F. Scott Fitzgerald. He said, the test

318
00:23:32,880 --> 00:23:36,840
Speaker 1: of a first rate intelligence is the ability to hold

319
00:23:37,320 --> 00:23:42,040
Speaker 1: two opposed ideas in the mind at the same time

320
00:23:42,920 --> 00:23:48,639
Speaker 1: and still retain the ability to function. One should, for example,

321
00:23:48,720 --> 00:23:52,240
Speaker 1: he said, be able to see that things are hopeless

322
00:23:53,160 --> 00:23:58,399
Speaker 1: and yet be determined to make them otherwise. That second

323
00:23:58,480 --> 00:24:02,440
Speaker 1: part is so needed right now, That's what our stories need,

324
00:24:03,400 --> 00:24:07,320
Speaker 1: accepting that things are tough, things are really hard, and

325
00:24:07,400 --> 00:24:10,800
Speaker 1: at the same time reminding each other that you can

326
00:24:10,920 --> 00:24:14,240
Speaker 1: make a change, You can transform your life, you can

327
00:24:14,320 --> 00:24:18,840
Speaker 1: take accountability, you can take action, you do have agency.

328
00:24:19,680 --> 00:24:24,159
Speaker 1: Reminding the world that extraordinary things have always been achieved

329
00:24:24,640 --> 00:24:27,879
Speaker 1: by a group of ordinary people. I'll leave you with this.

330
00:24:28,880 --> 00:24:32,200
Speaker 1: Imagine you walk into a party. At first, it looks fun,

331
00:24:32,480 --> 00:24:36,439
Speaker 1: people laughing, music playing, stories being told. But then you

332
00:24:36,520 --> 00:24:40,919
Speaker 1: notice something strange. Everywhere you turn, someone's doing better than you,

333
00:24:41,640 --> 00:24:45,880
Speaker 1: someone richer, someone prettier, someone with more friends, more followers,

334
00:24:45,920 --> 00:24:49,720
Speaker 1: more success. You walk into another room, and this one

335
00:24:49,720 --> 00:24:53,800
Speaker 1: feels worse. The room is full of arguments, everyone's shouting,

336
00:24:53,840 --> 00:24:57,639
Speaker 1: no one's listening, and the louder and angrier someone is

337
00:24:57,920 --> 00:25:01,720
Speaker 1: the bigger the crowd around them. That's when it hits you.

338
00:25:01,720 --> 00:25:04,400
Speaker 1: You never chose to come to this room. You were

339
00:25:04,440 --> 00:25:08,840
Speaker 1: invited by the algorithm. That's the cruel genius of social media.

340
00:25:09,520 --> 00:25:13,080
Speaker 1: It doesn't force us into comparison. It discovers we're already

341
00:25:13,160 --> 00:25:17,280
Speaker 1: drawn to it. It doesn't create division, It learns that

342
00:25:17,400 --> 00:25:21,840
Speaker 1: anger holds our gaze longer than joy. The algorithm didn't

343
00:25:21,880 --> 00:25:27,159
Speaker 1: create outrage. It turned outrage into entertainment. And here's the

344
00:25:27,240 --> 00:25:31,359
Speaker 1: question only you can answer. When you pick up your

345
00:25:31,359 --> 00:25:37,320
Speaker 1: phone tonight, are you walking back into that same party

346
00:25:37,480 --> 00:25:41,240
Speaker 1: or will you finally leave? Thank you for listening. I

347
00:25:41,280 --> 00:25:44,760
Speaker 1: hope you've subscribed. Share this episode with someone who needs

348
00:25:44,800 --> 00:25:47,919
Speaker 1: to hear it, And remember I'm forever in your corner

349
00:25:48,440 --> 00:25:51,560
Speaker 1: and I'm always rooting for you. If you love this episode,

350
00:25:51,680 --> 00:25:55,080
Speaker 1: you will also love my interview with Charles Douhig on

351
00:25:55,200 --> 00:25:59,080
Speaker 1: how to hack your brain, change any habit effortlessly, and

352
00:25:59,119 --> 00:26:02,879
Speaker 1: the secret to making better decisions. Look, am I hesitating

353
00:26:02,880 --> 00:26:05,359
Speaker 1: on this because I'm scared of making the choice, because

354
00:26:05,359 --> 00:26:07,560
Speaker 1: I'm scared of doing the work, Or am I sitting

355
00:26:07,560 --> 00:26:10,280
Speaker 1: with this because it just doesn't feel right yet