1 00:00:08,000 --> 00:00:10,320 Speaker 1: I know I need to get ready for work, but 2 00:00:10,400 --> 00:00:14,280 Speaker 1: I'm just so so tired. Maybe I can skip work 3 00:00:14,400 --> 00:00:19,880 Speaker 1: to take a nap justice once. What's that? Oh, my 4 00:00:19,960 --> 00:00:23,520 Speaker 1: depression monitors detecting early signs of a depressive state. I 5 00:00:23,560 --> 00:00:26,680 Speaker 1: thought it was just tired, but this might be a 6 00:00:26,720 --> 00:00:30,280 Speaker 1: bigger issue. Let me set in an appointment with my therapist. 7 00:00:40,960 --> 00:00:44,120 Speaker 1: Hey there, I'm grain class and this is technically speaking 8 00:00:44,360 --> 00:00:47,640 Speaker 1: an Intel podcast. The show is dedicated to highlighting the 9 00:00:47,680 --> 00:00:51,560 Speaker 1: way technology is revolutionizing the way we live, work and move. 10 00:00:52,479 --> 00:00:55,080 Speaker 1: In every episode, we'll connect with innovators in areas like 11 00:00:55,200 --> 00:00:59,440 Speaker 1: artificial intelligence to better understand the human centered technology they've developed. 12 00:01:00,600 --> 00:01:04,160 Speaker 1: Mental health care solutions remain underinvested in many communities around 13 00:01:04,160 --> 00:01:06,960 Speaker 1: the world, yet so many suffer from issues that they 14 00:01:06,959 --> 00:01:10,200 Speaker 1: don't even know. They have a lot of these conversations 15 00:01:10,240 --> 00:01:13,120 Speaker 1: around health care, hinge on making therapy more accessible to 16 00:01:13,160 --> 00:01:17,400 Speaker 1: those in need. However, it can be difficult to determine 17 00:01:17,440 --> 00:01:22,120 Speaker 1: that one is experiencing depression or mental health crisis. Artificial 18 00:01:22,160 --> 00:01:25,440 Speaker 1: intelligence is at the forefront of many different advancements in healthcare, 19 00:01:26,200 --> 00:01:28,400 Speaker 1: but today we are going to dive into how it 20 00:01:28,520 --> 00:01:32,280 Speaker 1: is working to make mental health care more accessible. To everyone. 21 00:01:33,080 --> 00:01:35,240 Speaker 1: In order to do that, I have to introduce you 22 00:01:35,400 --> 00:01:41,120 Speaker 1: to a special guest joining me now is Tinasawho. Tina 23 00:01:41,120 --> 00:01:44,000 Speaker 1: Saho was a high school student when she started exploring 24 00:01:44,040 --> 00:01:47,880 Speaker 1: coding and engineering with AI. She never considered herself to 25 00:01:47,920 --> 00:01:51,600 Speaker 1: be overly interested in technology to start. However, she took 26 00:01:51,600 --> 00:01:54,480 Speaker 1: her natural curiosity and eventually was invited to be a 27 00:01:54,480 --> 00:01:57,800 Speaker 1: part of Intel's AI for Youth Pilot program, where she 28 00:01:57,840 --> 00:02:01,200 Speaker 1: developed a tool that uses AI to detect and predict 29 00:02:01,240 --> 00:02:05,160 Speaker 1: patterns of depression with around eighty percent accuracy. Since then, 30 00:02:05,320 --> 00:02:08,240 Speaker 1: she's been awarded the Calm Trust Fellowship for Women. She 31 00:02:08,280 --> 00:02:11,720 Speaker 1: currently attends to Ayabata College at the University of Delhi, 32 00:02:12,160 --> 00:02:15,679 Speaker 1: where she's completing a bachelor's in computer science and exploring 33 00:02:15,720 --> 00:02:18,680 Speaker 1: more opportunities to use AI as a tool for mental 34 00:02:18,680 --> 00:02:19,480 Speaker 1: health and more. 35 00:02:19,960 --> 00:02:22,840 Speaker 2: Welcome to the show, Tina, Hi, thank you for inviting 36 00:02:22,880 --> 00:02:25,679 Speaker 2: me to this show, and I'm feeling more than privileged 37 00:02:25,720 --> 00:02:28,280 Speaker 2: to be a part of this podcast. So numbers stay 38 00:02:28,320 --> 00:02:28,880 Speaker 2: to everyone. 39 00:02:33,360 --> 00:02:36,040 Speaker 1: I'm really interested in your story of how you got 40 00:02:36,080 --> 00:02:40,200 Speaker 1: into the STEM field, what prompted your interest in that 41 00:02:40,280 --> 00:02:41,760 Speaker 1: field and also in AI. 42 00:02:42,360 --> 00:02:44,720 Speaker 2: To begin with, till my tenth class, I was a 43 00:02:44,760 --> 00:02:47,920 Speaker 2: student or a person who was always against technology, like 44 00:02:48,120 --> 00:02:50,519 Speaker 2: I always found in fact that the negative impacts that 45 00:02:50,600 --> 00:02:53,880 Speaker 2: technology brings and whatever ethical concerns are there, they cannot 46 00:02:53,880 --> 00:02:56,360 Speaker 2: be solved and they are simply too much to be 47 00:02:56,400 --> 00:03:00,000 Speaker 2: on the side of technology. But I remember in twenty nineteen, 48 00:03:00,400 --> 00:03:03,080 Speaker 2: the AAFO Youth program was launched and it was launched 49 00:03:03,120 --> 00:03:06,160 Speaker 2: in our school, which is Salvangal Senior Secondary School. So 50 00:03:06,600 --> 00:03:09,919 Speaker 2: when I participated in that program, I got to know 51 00:03:09,960 --> 00:03:14,720 Speaker 2: what artificial intelligence is, how many astonishing and stounding possibilities 52 00:03:14,760 --> 00:03:18,520 Speaker 2: AI can unlock and has already unlocked. And therefore, this 53 00:03:18,639 --> 00:03:22,600 Speaker 2: program pivoted my you know journey, my career, my professional journey, 54 00:03:22,800 --> 00:03:24,919 Speaker 2: and my cademic journey from being a non tech student 55 00:03:24,960 --> 00:03:28,160 Speaker 2: to becoming a tech student finally pursuing computer science as 56 00:03:28,320 --> 00:03:29,040 Speaker 2: my graduation. 57 00:03:29,960 --> 00:03:35,200 Speaker 1: Okay, was there a particular topic or person that really 58 00:03:35,240 --> 00:03:38,080 Speaker 1: did spark that interest? What was the particular topic that 59 00:03:38,160 --> 00:03:40,240 Speaker 1: actually got you really fired up? 60 00:03:40,880 --> 00:03:43,200 Speaker 2: So in my school there were two teachers who actually 61 00:03:43,280 --> 00:03:46,760 Speaker 2: you know, molded my thinking and they actually infused this 62 00:03:46,840 --> 00:03:49,640 Speaker 2: critical thinking aspect in me and they opened me to 63 00:03:49,680 --> 00:03:53,840 Speaker 2: the world of possibilities that science, that technology and engineering offers. 64 00:03:54,080 --> 00:03:55,800 Speaker 2: And I would like to take their name as well 65 00:03:55,800 --> 00:03:58,040 Speaker 2: as a you know, token of respect. So it's one 66 00:03:58,120 --> 00:04:02,080 Speaker 2: that Mam and Serbiam they empowered me. They encouraged me 67 00:04:02,160 --> 00:04:04,360 Speaker 2: to think beyond what I see. 68 00:04:04,920 --> 00:04:09,160 Speaker 1: And that led you to the Intel's AI for Youth program. 69 00:04:09,760 --> 00:04:12,640 Speaker 1: What sort of projects have you worked on? Are you 70 00:04:12,760 --> 00:04:13,800 Speaker 1: working on right now? 71 00:04:14,480 --> 00:04:17,760 Speaker 2: This isn't my participating in this program. I build Happiness Guru, 72 00:04:17,800 --> 00:04:21,120 Speaker 2: which is a model that predicts depression. Apart from that too, 73 00:04:21,240 --> 00:04:24,440 Speaker 2: I was also a part of Utul Tinkling Labs. Through 74 00:04:24,480 --> 00:04:27,560 Speaker 2: the Detail Lab of our school. I basically got to 75 00:04:27,600 --> 00:04:31,239 Speaker 2: know about Intel's AfOR Youth program only and in there itself, 76 00:04:31,320 --> 00:04:33,320 Speaker 2: I build a few projects and one of them was 77 00:04:33,360 --> 00:04:34,119 Speaker 2: Happiness Guru. 78 00:04:34,720 --> 00:04:38,720 Speaker 1: And with the Happiness Crew, is that related to the 79 00:04:38,760 --> 00:04:44,839 Speaker 1: depression detection research you've been doing? What kind of prompted 80 00:04:44,880 --> 00:04:47,200 Speaker 1: you to look in that direction in the field of 81 00:04:47,880 --> 00:04:51,840 Speaker 1: depression and then using technology? Because generally, speaker, we don't 82 00:04:51,880 --> 00:04:57,320 Speaker 1: associate technology with treating depression or even detecting depression. So 83 00:04:57,400 --> 00:04:59,280 Speaker 1: what was the spark for you there? 84 00:05:00,240 --> 00:05:02,440 Speaker 2: So while I was in this program, I was in 85 00:05:02,480 --> 00:05:05,600 Speaker 2: Class ten and my Plus ten results were out and 86 00:05:05,680 --> 00:05:08,240 Speaker 2: they were not as much as I expected, and I 87 00:05:08,400 --> 00:05:12,120 Speaker 2: went into a phase of depression because I associated myself 88 00:05:12,160 --> 00:05:15,640 Speaker 2: worth with the Marxi score, so my own personal experience 89 00:05:15,680 --> 00:05:19,320 Speaker 2: of dealing with depression. And then at that time, you 90 00:05:19,360 --> 00:05:21,919 Speaker 2: know that society rates were very alarming among the youth, 91 00:05:22,040 --> 00:05:25,440 Speaker 2: especially aged between fifteen to twenty nine. And we found 92 00:05:25,480 --> 00:05:28,600 Speaker 2: that the driving forces behind you know these when these 93 00:05:28,640 --> 00:05:32,799 Speaker 2: societies were you know, peer pressure, overburdening academics, financial stress, 94 00:05:33,320 --> 00:05:36,039 Speaker 2: and too much expectations that we have from youth, you know, 95 00:05:36,120 --> 00:05:38,839 Speaker 2: especially if you talk about teenage and someone who is 96 00:05:39,279 --> 00:05:42,320 Speaker 2: in between eighteen to twenty five. So on researching, we 97 00:05:42,360 --> 00:05:45,440 Speaker 2: found that these societ rates are very alarming, they're very distressing, 98 00:05:45,480 --> 00:05:48,240 Speaker 2: and these are the leading cause of the depression or stress. 99 00:05:48,600 --> 00:05:51,159 Speaker 2: And thereby we thought that we must come up with 100 00:05:51,200 --> 00:05:54,880 Speaker 2: some solution that can basically help us predict which person 101 00:05:54,920 --> 00:05:56,880 Speaker 2: is going through depression, and that to in a very 102 00:05:56,920 --> 00:06:00,640 Speaker 2: human friendly manner, not making someone uncomfortable with the kind 103 00:06:00,680 --> 00:06:02,560 Speaker 2: of procedures or with the kind of system we have. 104 00:06:03,160 --> 00:06:07,000 Speaker 2: So these were the I would say, the enablers that 105 00:06:07,120 --> 00:06:09,320 Speaker 2: led our team building this solution. 106 00:06:09,920 --> 00:06:13,440 Speaker 1: And in terms of the happiness grew app can you 107 00:06:13,520 --> 00:06:16,120 Speaker 1: just explain how it actually works, you know, to try 108 00:06:16,120 --> 00:06:18,480 Speaker 1: and detect the early signs of depression. 109 00:06:19,279 --> 00:06:23,160 Speaker 2: First of all, it's our web based application. While building 110 00:06:23,200 --> 00:06:26,080 Speaker 2: this project, the queue that we took, you know, to 111 00:06:26,160 --> 00:06:29,080 Speaker 2: build the entire model was that after our research, we 112 00:06:29,160 --> 00:06:32,160 Speaker 2: got to know that a person's vocabulary can be a 113 00:06:32,200 --> 00:06:35,800 Speaker 2: mirror into their mental state. And taking this as the queue, 114 00:06:36,080 --> 00:06:39,640 Speaker 2: we build this project which tries to analyze the emotional 115 00:06:39,760 --> 00:06:42,960 Speaker 2: quotient of a person of a user through their facial 116 00:06:43,000 --> 00:06:46,640 Speaker 2: expression and then their textual responses that the user is 117 00:06:46,640 --> 00:06:49,120 Speaker 2: going to provide to the AA machine. So the working 118 00:06:49,160 --> 00:06:51,680 Speaker 2: of the project is divided into three steps. The first 119 00:06:51,720 --> 00:06:54,919 Speaker 2: step is emotion detection stage, and in this stage you 120 00:06:55,040 --> 00:06:57,760 Speaker 2: basically need to stand in front of your laptop or 121 00:06:57,800 --> 00:07:00,720 Speaker 2: whatever device you are using this web application, and then 122 00:07:00,760 --> 00:07:03,200 Speaker 2: it detects your current mode, whether you're happy or sad, 123 00:07:03,200 --> 00:07:06,279 Speaker 2: you're neutral, angry. Then the next step is that user 124 00:07:06,360 --> 00:07:10,280 Speaker 2: is asked to answer nine questions and there's a scale 125 00:07:10,280 --> 00:07:13,280 Speaker 2: of relevance and then they need to select how much 126 00:07:13,320 --> 00:07:15,560 Speaker 2: relevant or how much they are able to relate this 127 00:07:15,800 --> 00:07:19,000 Speaker 2: to this situation. Then, after these two steps, a threshold 128 00:07:19,080 --> 00:07:21,720 Speaker 2: score is generated which gives the initial lead. If the 129 00:07:21,760 --> 00:07:23,960 Speaker 2: person is stressed or not, and if the score is 130 00:07:23,960 --> 00:07:26,240 Speaker 2: below the threshold that we have said, the person is 131 00:07:26,280 --> 00:07:29,320 Speaker 2: predicted as happy, while in the other case, user is 132 00:07:29,360 --> 00:07:31,200 Speaker 2: taken to the third step, which is the final step. 133 00:07:31,400 --> 00:07:34,160 Speaker 2: And this step consists of four descriptive questions which he 134 00:07:34,320 --> 00:07:36,960 Speaker 2: or she can use as a platform to went out 135 00:07:37,080 --> 00:07:40,600 Speaker 2: all his or her feelings and thoughts. So whatever answers 136 00:07:40,720 --> 00:07:44,200 Speaker 2: user will give to these four descriptive questions, these answers 137 00:07:44,200 --> 00:07:47,200 Speaker 2: will be used as a basis of classification. Then the 138 00:07:47,240 --> 00:07:49,560 Speaker 2: machine will predict whether the user is depressed or not. 139 00:07:49,680 --> 00:07:52,480 Speaker 2: So this AI machine, whatever you know input we are 140 00:07:52,480 --> 00:07:55,360 Speaker 2: giving in this step. There's a model namely SVM, which 141 00:07:55,360 --> 00:07:58,760 Speaker 2: is support vector machine. It's a non contextual classification model. 142 00:07:58,760 --> 00:08:02,120 Speaker 2: It is basically used to classify things. And then we 143 00:08:02,240 --> 00:08:04,360 Speaker 2: are using this model on the kind of you know, 144 00:08:04,480 --> 00:08:07,040 Speaker 2: language or keywords that are used in the answers. And 145 00:08:07,080 --> 00:08:09,760 Speaker 2: then accordingly the results are given out that whether the 146 00:08:09,760 --> 00:08:12,000 Speaker 2: person is stressed or not, and if the person is stressed, 147 00:08:12,480 --> 00:08:16,520 Speaker 2: automatically the person is consulted to the concerned authorities or 148 00:08:16,560 --> 00:08:20,600 Speaker 2: counselor otherwise the person is predicted as happy or not stressed. 149 00:08:22,440 --> 00:08:26,320 Speaker 1: Detecting and treating mental health is something with which many 150 00:08:26,400 --> 00:08:30,640 Speaker 1: societies around the world struggle. According to the World Health Organization, 151 00:08:31,160 --> 00:08:33,880 Speaker 1: approximately two hundred and eighty million people in the world 152 00:08:33,920 --> 00:08:37,160 Speaker 1: suffer from depression and more than three hundred million are 153 00:08:37,200 --> 00:08:41,479 Speaker 1: living with anxiety. Many people with these mental health conditions 154 00:08:41,559 --> 00:08:46,000 Speaker 1: exhibit some symptoms as children or young adults, but based 155 00:08:46,040 --> 00:08:49,000 Speaker 1: on guidance from the US National Institute of Mental Health, 156 00:08:49,800 --> 00:08:53,560 Speaker 1: depression can only be diagnosed once an individual exhibits the 157 00:08:53,640 --> 00:08:57,400 Speaker 1: five major symptoms of depression every day, all day for 158 00:08:57,440 --> 00:09:00,680 Speaker 1: a minimum of two weeks. Imagine how we could help 159 00:09:00,720 --> 00:09:04,280 Speaker 1: people earlier if we were able to identify depression with 160 00:09:04,400 --> 00:09:07,200 Speaker 1: the help of AI tools like the Happiness Guru model. 161 00:09:10,120 --> 00:09:14,600 Speaker 1: How does one actually create that model? What data is 162 00:09:14,760 --> 00:09:18,760 Speaker 1: needed to train that model so that it can get 163 00:09:18,760 --> 00:09:19,440 Speaker 1: that output. 164 00:09:20,040 --> 00:09:23,120 Speaker 2: So basically, whenever we build any project, we were taught 165 00:09:23,120 --> 00:09:25,120 Speaker 2: this thing in the program itself that there's a whole 166 00:09:25,120 --> 00:09:27,840 Speaker 2: project cycle that needs to be taken into you know, 167 00:09:27,840 --> 00:09:30,959 Speaker 2: account while we're building any project. So the first step 168 00:09:31,040 --> 00:09:33,800 Speaker 2: that comes into the AA project cycle is problem scoping. 169 00:09:33,840 --> 00:09:36,560 Speaker 2: So we have problem statements, we have a stakeholders, and 170 00:09:36,760 --> 00:09:39,040 Speaker 2: we have our ideal solution as well. Now comes data 171 00:09:39,080 --> 00:09:42,600 Speaker 2: acquisition so basically to make this project work the way 172 00:09:42,640 --> 00:09:45,280 Speaker 2: it is working right now, data was collected you know, 173 00:09:45,320 --> 00:09:48,679 Speaker 2: anonymously through offline and online surveys and across five different 174 00:09:48,679 --> 00:09:52,680 Speaker 2: schools across India. So during these surveys, we briefed students 175 00:09:52,720 --> 00:09:55,320 Speaker 2: in the school what this survey is about and then 176 00:09:55,360 --> 00:09:58,040 Speaker 2: they were asked to fill out that form which contained 177 00:09:58,120 --> 00:10:01,400 Speaker 2: descriptive questions. Now, these descriptive questions that we selected, these 178 00:10:01,440 --> 00:10:04,679 Speaker 2: were validated by a team of psychiatrists and counselors and 179 00:10:04,720 --> 00:10:07,160 Speaker 2: then with the help of this survey process, we were 180 00:10:07,200 --> 00:10:09,640 Speaker 2: able to develop an authentic data set of seven hundred 181 00:10:09,679 --> 00:10:13,880 Speaker 2: plus centuries where the students basically wrote whatever they felt 182 00:10:13,920 --> 00:10:16,400 Speaker 2: during that time and you know, went out their thoughts 183 00:10:16,480 --> 00:10:19,360 Speaker 2: in that survey. The responses were labeled the on the 184 00:10:19,360 --> 00:10:21,880 Speaker 2: scale of A two D, with A being least sever 185 00:10:22,120 --> 00:10:26,240 Speaker 2: like perfectly healthy mentally and to D being needing immediate 186 00:10:26,280 --> 00:10:29,040 Speaker 2: support from professionals and family. And this was done with 187 00:10:29,080 --> 00:10:32,000 Speaker 2: the help of our school counselor, Ishitan Atara, So she 188 00:10:32,080 --> 00:10:35,400 Speaker 2: helped us in you know, laboring these responses and then 189 00:10:35,480 --> 00:10:38,880 Speaker 2: this data was used to train that SVM model that 190 00:10:38,960 --> 00:10:41,280 Speaker 2: I was talking about that is a part of step three, 191 00:10:41,679 --> 00:10:44,719 Speaker 2: So we need to convert this offline data into a 192 00:10:44,760 --> 00:10:48,439 Speaker 2: digitized format because that's how model gets trained. So we 193 00:10:48,480 --> 00:10:51,280 Speaker 2: did that, we started classifying, and then we trained the 194 00:10:51,400 --> 00:10:54,280 Speaker 2: SVM model. Apart from that, there's one more thing that 195 00:10:54,320 --> 00:10:56,679 Speaker 2: has went into this. The step one which I talked 196 00:10:56,679 --> 00:11:00,560 Speaker 2: about is about, you know, recognizing whatever current emotion the 197 00:11:00,640 --> 00:11:03,760 Speaker 2: user has, whatever their emotion is currently while they're using 198 00:11:03,840 --> 00:11:07,679 Speaker 2: So this is turned with the help of library basically 199 00:11:07,720 --> 00:11:10,480 Speaker 2: fast a dot Vision. So fast a dot Vision is 200 00:11:10,480 --> 00:11:13,280 Speaker 2: a library that is used for computer vision tasks. And 201 00:11:13,320 --> 00:11:15,720 Speaker 2: then we have trained this module using a data set. 202 00:11:15,880 --> 00:11:18,520 Speaker 2: So this data set consisted of two thousand rows i 203 00:11:18,520 --> 00:11:22,120 Speaker 2: would say, which consisted of facial expressions of different people, 204 00:11:22,240 --> 00:11:25,680 Speaker 2: like there were videos and images of people from different 205 00:11:26,200 --> 00:11:30,040 Speaker 2: genders and heritages of different backgrounds, and then they were 206 00:11:30,120 --> 00:11:33,360 Speaker 2: classified as happy saturn nedle to train our module, which 207 00:11:33,440 --> 00:11:34,680 Speaker 2: was fast a Dot Vision. 208 00:11:36,840 --> 00:11:40,120 Speaker 1: What Tina is describing in her design philosophy is very 209 00:11:40,200 --> 00:11:44,000 Speaker 1: interesting because in a way it mirrors processes used by 210 00:11:44,080 --> 00:11:49,720 Speaker 1: psychiatrists and counselors to identify depression in young people at schools. However, 211 00:11:50,000 --> 00:11:55,319 Speaker 1: in her system the effectiveness is amplified. Oftentimes people experiencing 212 00:11:55,360 --> 00:11:58,840 Speaker 1: depression are not able to recognize the symptoms in themselves, 213 00:11:59,120 --> 00:12:02,520 Speaker 1: and for young people particular, having access to a professional 214 00:12:02,520 --> 00:12:06,080 Speaker 1: who could observe and identify the science is not guaranteed. 215 00:12:06,720 --> 00:12:11,080 Speaker 1: For cultural, social and economic reasons, mental health is largely ignored. 216 00:12:11,720 --> 00:12:14,280 Speaker 1: I can see the benefit of an automated system being 217 00:12:14,360 --> 00:12:17,080 Speaker 1: used to identify it and how that can help those 218 00:12:17,240 --> 00:12:21,000 Speaker 1: with reservations around mental healthcare take that crucial first step. 219 00:12:22,880 --> 00:12:26,440 Speaker 1: You're listening to technically Speaking an Intel podcast will be 220 00:12:26,520 --> 00:12:36,679 Speaker 1: right back. Welcome back to technically Speaking an Intel podcast. 221 00:12:37,000 --> 00:12:42,160 Speaker 1: I'm here now with Tina. So, So, in terms of 222 00:12:42,200 --> 00:12:46,920 Speaker 1: your research or next phase, do you think these sorts 223 00:12:47,320 --> 00:12:52,720 Speaker 1: of wearable devices or things that can detect people's emotions, 224 00:12:53,200 --> 00:12:55,360 Speaker 1: do you see a future where that could be a 225 00:12:55,400 --> 00:12:58,920 Speaker 1: possibility where we could get in early in terms of 226 00:12:59,280 --> 00:13:00,720 Speaker 1: detecting depression. 227 00:13:01,480 --> 00:13:03,600 Speaker 2: Yes, there can be. In fact, there's been a rise 228 00:13:03,640 --> 00:13:06,160 Speaker 2: in it lately. Like I've been following up the news 229 00:13:06,200 --> 00:13:09,160 Speaker 2: around this, and I've got to know that there was 230 00:13:09,200 --> 00:13:12,520 Speaker 2: some institute in New York itself which conducted a study 231 00:13:12,520 --> 00:13:15,560 Speaker 2: which basically built an machine learning model that took the 232 00:13:15,679 --> 00:13:18,840 Speaker 2: data of thousands users, and then this model was able 233 00:13:18,880 --> 00:13:21,760 Speaker 2: to tell whether a person was mentally healthy or not. 234 00:13:22,280 --> 00:13:25,080 Speaker 2: So we need to understand how this works for us 235 00:13:25,120 --> 00:13:28,160 Speaker 2: to fall like, we are basically collecting data points in 236 00:13:28,280 --> 00:13:30,800 Speaker 2: terms of different variables, and these variables are like you know, 237 00:13:31,080 --> 00:13:33,200 Speaker 2: what at our pulse rate, what is our heart beat? 238 00:13:33,679 --> 00:13:36,200 Speaker 2: And I mean different things that can be measured by 239 00:13:36,240 --> 00:13:39,480 Speaker 2: these devices, by these variables to find the relation between 240 00:13:39,640 --> 00:13:42,600 Speaker 2: someone's mental health and whatever data points we are collecting. 241 00:13:43,040 --> 00:13:46,280 Speaker 2: So there's a possibility that in the coming year we 242 00:13:46,480 --> 00:13:49,040 Speaker 2: can lead mental health care services. Apart from this, a 243 00:13:49,080 --> 00:13:50,920 Speaker 2: similar thing that strucks to me right now is that 244 00:13:51,120 --> 00:13:55,359 Speaker 2: brain computer interface. I mean well, brain computer interface is 245 00:13:55,400 --> 00:13:58,680 Speaker 2: a machine that actually helps us to control a device 246 00:13:58,720 --> 00:14:02,480 Speaker 2: or machine using our brain. So if something of that 247 00:14:02,559 --> 00:14:05,640 Speaker 2: sort can be infused with machine learning, and then if 248 00:14:05,640 --> 00:14:08,800 Speaker 2: we can build some solution that is oriented towards solving 249 00:14:08,880 --> 00:14:11,720 Speaker 2: mental health problems that exist, that is oriented towards providing 250 00:14:11,720 --> 00:14:15,120 Speaker 2: more healthcare services, like those that accessible enough and affordable 251 00:14:15,440 --> 00:14:18,280 Speaker 2: as well, So I think majority of problems can be 252 00:14:18,360 --> 00:14:21,680 Speaker 2: solved in this area. 253 00:14:21,880 --> 00:14:26,600 Speaker 1: Tina mentioning BCI or brain computer interface reminds me of 254 00:14:26,600 --> 00:14:30,840 Speaker 1: the conversation in episode three with Jaggedish and Lama. We 255 00:14:30,920 --> 00:14:34,720 Speaker 1: tend to think of BCI as human brains controlling the 256 00:14:34,760 --> 00:14:37,920 Speaker 1: function of a machine, like moving a mouse cursor or 257 00:14:37,960 --> 00:14:42,000 Speaker 1: controlling a robotic limb. However, Tina imagines a world where 258 00:14:42,040 --> 00:14:45,280 Speaker 1: our brains can simply inform machines on how to service us. 259 00:14:46,080 --> 00:14:47,880 Speaker 1: It is not so much that you would need to 260 00:14:48,040 --> 00:14:51,200 Speaker 1: even think about being helped, but the machine learning process 261 00:14:51,200 --> 00:14:53,560 Speaker 1: would allow a tool to remind you of a service 262 00:14:53,600 --> 00:14:56,920 Speaker 1: you need. It's almost like having a second brain. I 263 00:14:56,960 --> 00:14:59,640 Speaker 1: can't wait to see all of the medical applications this 264 00:14:59,720 --> 00:15:05,160 Speaker 1: open up to the world. Particularly through the pandemic and 265 00:15:05,200 --> 00:15:09,160 Speaker 1: post pandemic, there was a rise in mental health issues 266 00:15:09,200 --> 00:15:14,080 Speaker 1: which needed expert care. Now do you think that AI 267 00:15:14,400 --> 00:15:18,680 Speaker 1: can play a role in actually providing therapy for people 268 00:15:18,720 --> 00:15:23,040 Speaker 1: with mental health concerns. I recently read an article in 269 00:15:23,160 --> 00:15:27,920 Speaker 1: Time magazine about robot, which is a AI personal therapist. 270 00:15:28,680 --> 00:15:31,720 Speaker 1: I'd like to get your thoughts as to whether they 271 00:15:31,760 --> 00:15:36,240 Speaker 1: could actually provide useful advice for people to help manage 272 00:15:36,320 --> 00:15:38,440 Speaker 1: their depression and mental health issues. 273 00:15:39,400 --> 00:15:41,680 Speaker 2: So, when we think of mental health care, you know, 274 00:15:41,960 --> 00:15:45,960 Speaker 2: the corner store of this is communication. It's not depending 275 00:15:46,040 --> 00:15:49,400 Speaker 2: on the procedures, but more on the communication. Like if 276 00:15:49,480 --> 00:15:52,000 Speaker 2: we know that therapist and the patient that there should 277 00:15:52,000 --> 00:15:54,840 Speaker 2: be a strong relationship between them. The relationship should be 278 00:15:54,840 --> 00:15:57,520 Speaker 2: good enough so that the patient can communicate with their 279 00:15:57,560 --> 00:16:01,160 Speaker 2: therapists and then the problem can whatever problem the patient 280 00:16:01,200 --> 00:16:05,040 Speaker 2: is going through. So like if you talk about therapists 281 00:16:05,040 --> 00:16:07,520 Speaker 2: in terms of air, So there are chadbots which are 282 00:16:07,520 --> 00:16:10,960 Speaker 2: coming up, like robot and new par So these chadbots 283 00:16:11,000 --> 00:16:14,400 Speaker 2: that are increasingly being used to offer advice and a 284 00:16:14,440 --> 00:16:17,720 Speaker 2: line of communication for mental health patients during their treatment. 285 00:16:18,120 --> 00:16:20,520 Speaker 2: So they can also help with coping up with symptoms 286 00:16:20,560 --> 00:16:22,520 Speaker 2: as well as they can look out for keyword that 287 00:16:22,560 --> 00:16:26,520 Speaker 2: could trigger a possible help that patient needs. So chargipity 288 00:16:26,600 --> 00:16:28,560 Speaker 2: can be used like a therapist. Like there have been 289 00:16:28,640 --> 00:16:31,280 Speaker 2: certain use cases, like I've been reading on Reddit and 290 00:16:31,320 --> 00:16:33,280 Speaker 2: there have been people who have been like sharing their 291 00:16:33,320 --> 00:16:36,520 Speaker 2: stories around how they use chargity as a therapist. So 292 00:16:36,760 --> 00:16:39,440 Speaker 2: when we see that chatbot can be used as a therapist, 293 00:16:39,720 --> 00:16:42,440 Speaker 2: it is like we are giving them some inputs and 294 00:16:42,440 --> 00:16:45,080 Speaker 2: they're basically doing sentiment analygies on the basis of textual 295 00:16:45,120 --> 00:16:47,440 Speaker 2: responses that we're giving to them, and then they are 296 00:16:47,440 --> 00:16:50,600 Speaker 2: basically modifying their answers to make it more human like 297 00:16:51,040 --> 00:16:54,880 Speaker 2: and that's how they can work as AI therapist. But 298 00:16:55,080 --> 00:16:58,040 Speaker 2: there are concerns around it as well. Like the first 299 00:16:58,080 --> 00:17:01,160 Speaker 2: thing that comes up with is reliability. How much accurate 300 00:17:01,200 --> 00:17:04,600 Speaker 2: of the solution that chatbot is providing us or any 301 00:17:04,640 --> 00:17:06,920 Speaker 2: tool that we have built as a form of therapist 302 00:17:07,000 --> 00:17:11,360 Speaker 2: is providing us. So first is reliability and then comes accountability. 303 00:17:11,400 --> 00:17:14,399 Speaker 2: What if you know, something wrong happens, Who's responsible for 304 00:17:14,480 --> 00:17:17,679 Speaker 2: all of this? But apart from this, the concern that 305 00:17:17,800 --> 00:17:20,359 Speaker 2: always struck me is that these are privately funded apps, 306 00:17:20,400 --> 00:17:23,359 Speaker 2: Like these are the apps that have been used at 307 00:17:23,359 --> 00:17:26,760 Speaker 2: commercial level. I mean, there are certain subscription charges that 308 00:17:26,800 --> 00:17:30,120 Speaker 2: need to be paid to use these apps. So I've 309 00:17:30,160 --> 00:17:33,080 Speaker 2: always had this view that once you start commercializing and 310 00:17:33,119 --> 00:17:37,840 Speaker 2: start making out profits from healthcare services, then things turn problematic, 311 00:17:37,920 --> 00:17:41,200 Speaker 2: you know, and when something as vulnerable and as volatile 312 00:17:41,240 --> 00:17:44,199 Speaker 2: as mental health is involved, I think we must be 313 00:17:44,359 --> 00:17:47,280 Speaker 2: very much cautious. We must be very much vigilant about 314 00:17:47,320 --> 00:17:49,160 Speaker 2: the kind of apps we are using and the kind 315 00:17:49,200 --> 00:17:51,800 Speaker 2: of tools that are coming in in terms of mental 316 00:17:51,840 --> 00:17:52,879 Speaker 2: health care services. 317 00:17:53,640 --> 00:17:56,000 Speaker 1: And that leads me to if you are going to 318 00:17:56,000 --> 00:17:58,800 Speaker 1: be using these sorts of chat pots like chat GPT, 319 00:17:58,960 --> 00:18:01,200 Speaker 1: as you mentioned, to make sure that you're well aware 320 00:18:01,200 --> 00:18:04,520 Speaker 1: of who's got your data, what the privacy concerns may be, 321 00:18:05,280 --> 00:18:07,919 Speaker 1: and how you can make an informed decision. I like 322 00:18:07,960 --> 00:18:11,119 Speaker 1: to get your thoughts around that, particularly around privacy and 323 00:18:11,200 --> 00:18:14,320 Speaker 1: data security, and maybe you could start with how you 324 00:18:14,400 --> 00:18:15,760 Speaker 1: tackled it with your app. 325 00:18:16,800 --> 00:18:18,960 Speaker 2: This is one of the main concerns that come up. 326 00:18:19,000 --> 00:18:22,120 Speaker 2: Like you also mentioned that whenever we are using such apps, 327 00:18:22,160 --> 00:18:23,720 Speaker 2: we need to be aware that what kind of data 328 00:18:23,760 --> 00:18:26,520 Speaker 2: we are feeding into it and what kind of formissions 329 00:18:26,520 --> 00:18:29,359 Speaker 2: we're giving to such a tool. But someone who is 330 00:18:29,520 --> 00:18:32,840 Speaker 2: going through a mental health problem mental illness, I mean 331 00:18:33,040 --> 00:18:36,480 Speaker 2: we cannot say that the person is healthy enough or 332 00:18:36,520 --> 00:18:38,840 Speaker 2: stable enough to be able to make a decision on this, 333 00:18:39,240 --> 00:18:42,840 Speaker 2: and therefore privacy concerns will come later in the stage. 334 00:18:42,920 --> 00:18:44,960 Speaker 2: But the first thing is that are we able to 335 00:18:45,520 --> 00:18:49,000 Speaker 2: make the patients familiarize with the kind of data they're 336 00:18:49,000 --> 00:18:51,840 Speaker 2: feeding into the apps and what are the consequences or 337 00:18:51,920 --> 00:18:53,919 Speaker 2: ramifications that this data can lead to. 338 00:18:54,520 --> 00:18:59,080 Speaker 1: Yeah, because I actually heard some stories around people using 339 00:18:59,119 --> 00:19:05,400 Speaker 1: these chat says therapy and the concept of this transference, 340 00:19:05,440 --> 00:19:09,840 Speaker 1: so they're actually falling in love with the bots. There's 341 00:19:09,880 --> 00:19:13,720 Speaker 1: a similar experience with psychologists where patients fall in love 342 00:19:13,760 --> 00:19:17,639 Speaker 1: with the therapist. So that's just another potential challenge that 343 00:19:17,680 --> 00:19:19,360 Speaker 1: we all have to come to deal with if you're 344 00:19:19,359 --> 00:19:20,719 Speaker 1: going to start using these things. 345 00:19:21,320 --> 00:19:24,520 Speaker 2: Yeah, they're a virtual entities that are coming into this 346 00:19:24,600 --> 00:19:26,920 Speaker 2: scenario and we are able to see them and they've 347 00:19:26,920 --> 00:19:30,320 Speaker 2: been living their own life. People are becoming so comfortable 348 00:19:30,359 --> 00:19:33,679 Speaker 2: with chatbirds now because definitely there's a lack of communication 349 00:19:33,720 --> 00:19:35,920 Speaker 2: that is happening, and ever since the pandemic gets strung, 350 00:19:36,000 --> 00:19:39,199 Speaker 2: this communication gap has increased, it has profoundly increased, so 351 00:19:39,240 --> 00:19:41,720 Speaker 2: people are finding way to escape this and then these 352 00:19:41,760 --> 00:19:44,639 Speaker 2: AI therapists come as a rescue and therefore people use 353 00:19:44,680 --> 00:19:48,560 Speaker 2: it blindly without being enough aware about what kind of 354 00:19:48,640 --> 00:19:51,080 Speaker 2: data they're feeling it and what kind of algorithms these 355 00:19:51,280 --> 00:19:54,760 Speaker 2: applications are using. Because we know that to these algorithms 356 00:19:54,880 --> 00:19:58,320 Speaker 2: may not be explainable, they're not transparent, so we have 357 00:19:58,400 --> 00:20:01,399 Speaker 2: to be aware about this as well. Literacy and education 358 00:20:01,520 --> 00:20:03,320 Speaker 2: is needed in these aspects as well. 359 00:20:04,040 --> 00:20:08,639 Speaker 1: Yeah, just on that you talked about explainability and transparency, 360 00:20:09,240 --> 00:20:11,040 Speaker 1: do we just explain to the audience who may be 361 00:20:11,119 --> 00:20:14,040 Speaker 1: not so familiar with those terms when it comes to 362 00:20:14,640 --> 00:20:18,359 Speaker 1: AI models, what that actually means transparency. 363 00:20:18,400 --> 00:20:20,840 Speaker 2: Okay, So there's a term that goes with algorithms, and 364 00:20:20,880 --> 00:20:24,320 Speaker 2: that's black box. So algorithms are like black blocks. We 365 00:20:24,400 --> 00:20:26,679 Speaker 2: know what is going out, but we do not know how 366 00:20:26,800 --> 00:20:31,240 Speaker 2: all of this is functioning, what is actually into the algorithm, 367 00:20:31,520 --> 00:20:33,959 Speaker 2: and what is the procedure and how on what basis 368 00:20:33,960 --> 00:20:37,840 Speaker 2: they're doing everything. Transparency is related to the kind of 369 00:20:37,920 --> 00:20:40,160 Speaker 2: data we're feeding it and the way we are using 370 00:20:40,200 --> 00:20:43,359 Speaker 2: it and how algorithm is working. To know this and 371 00:20:43,440 --> 00:20:47,520 Speaker 2: explainability means that any user, because there are two categories 372 00:20:47,800 --> 00:20:51,280 Speaker 2: of population who are associated with any AA system. The 373 00:20:51,359 --> 00:20:53,359 Speaker 2: first one are users and the second one are the 374 00:20:53,560 --> 00:20:56,639 Speaker 2: developers and stakeholders. So stakeholders must know that what kind 375 00:20:56,680 --> 00:20:59,000 Speaker 2: of algorithm it is and there should be transparency in it. 376 00:20:59,160 --> 00:21:01,679 Speaker 2: But when it comes to you user, AI systems and 377 00:21:01,720 --> 00:21:04,880 Speaker 2: those algorithms must be explainable enough. I mean users are 378 00:21:04,880 --> 00:21:08,160 Speaker 2: able to understand in a very human like language, that's 379 00:21:08,240 --> 00:21:09,960 Speaker 2: what this algorithm is doing. 380 00:21:10,520 --> 00:21:14,880 Speaker 1: That's really good And as AI emerges as this tool 381 00:21:14,920 --> 00:21:18,000 Speaker 1: to help people struggling with their mental health, I'd like 382 00:21:18,040 --> 00:21:20,720 Speaker 1: a few more comments just around how you see it 383 00:21:20,800 --> 00:21:24,280 Speaker 1: working in tandem with the medical community to better serve 384 00:21:24,800 --> 00:21:28,160 Speaker 1: their patients and their communities. Do you have any thoughts 385 00:21:28,240 --> 00:21:30,719 Speaker 1: on how you know this tool can actually be used 386 00:21:31,359 --> 00:21:33,520 Speaker 1: together rather than a replacement. 387 00:21:34,200 --> 00:21:38,280 Speaker 2: Yeah, Basically, we always think that AI is a disruptor. 388 00:21:38,359 --> 00:21:41,600 Speaker 2: We have always thought of this any technology that comes, 389 00:21:41,760 --> 00:21:44,040 Speaker 2: but I've always believed that they are over here to 390 00:21:44,119 --> 00:21:49,000 Speaker 2: augment our capabilities and to supplement whatever you know, roles 391 00:21:49,000 --> 00:21:52,600 Speaker 2: are there. So I'm from India and the very first 392 00:21:52,640 --> 00:21:54,159 Speaker 2: thing that I mean I have to cover up is 393 00:21:54,160 --> 00:21:57,560 Speaker 2: that we need to educate people around mental health because 394 00:21:57,560 --> 00:22:01,600 Speaker 2: in India, the most instrumental impedt in terms of mental 395 00:22:01,640 --> 00:22:03,960 Speaker 2: health is lack of awareness and education. People do not 396 00:22:04,080 --> 00:22:06,760 Speaker 2: know what exactly depression is, what exactly anxiety and stresses. 397 00:22:06,800 --> 00:22:09,320 Speaker 2: They use it in a very casual way. And to 398 00:22:09,359 --> 00:22:13,120 Speaker 2: be very honest, mental health is something which is stigmatized 399 00:22:13,119 --> 00:22:15,280 Speaker 2: in India. So you know, if someone is suffering from 400 00:22:15,280 --> 00:22:17,600 Speaker 2: mental health issue, they are often labeled as lunatics or 401 00:22:17,600 --> 00:22:20,840 Speaker 2: crazy or possessed. So we need to educate people around 402 00:22:20,840 --> 00:22:24,440 Speaker 2: this first of all. So I believe my project it's 403 00:22:24,480 --> 00:22:27,480 Speaker 2: still it's working. I'm looking forward to deploying it into 404 00:22:27,600 --> 00:22:30,720 Speaker 2: as many schools as I can. Because we know that annealgorithm, 405 00:22:30,800 --> 00:22:33,080 Speaker 2: the more data we feed into it, the more accurate 406 00:22:33,160 --> 00:22:36,480 Speaker 2: it becomes. Its current accuracy is seventy seven to eighty person. 407 00:22:36,880 --> 00:22:39,000 Speaker 2: So we need to increase that accuracy first of all, 408 00:22:39,280 --> 00:22:41,840 Speaker 2: and then we have to take care of the data. 409 00:22:42,160 --> 00:22:45,679 Speaker 2: We need to have some regulations, we have some norms 410 00:22:45,680 --> 00:22:48,439 Speaker 2: and rules. We have to inform our users also that 411 00:22:48,480 --> 00:22:51,639 Speaker 2: the data that we're taking from them is in safe hans. Secondly, 412 00:22:52,000 --> 00:22:55,520 Speaker 2: I believe I will be changing the working of this project. 413 00:22:55,680 --> 00:22:59,160 Speaker 2: Currently it works on you know, facial recognition on current mood, 414 00:22:59,240 --> 00:23:01,720 Speaker 2: and that can easily be fabricated. I mean something that 415 00:23:01,800 --> 00:23:04,040 Speaker 2: is not reliable. That is not a thing that should 416 00:23:04,080 --> 00:23:06,960 Speaker 2: be taken into account while you are assessing someone's mental health. 417 00:23:07,280 --> 00:23:09,359 Speaker 2: So I think I need to eliminate this step and 418 00:23:09,400 --> 00:23:12,640 Speaker 2: replace it with something better. It could possibly be like 419 00:23:13,200 --> 00:23:16,640 Speaker 2: I find a BCI like brain computer interface, this technology. 420 00:23:16,640 --> 00:23:19,280 Speaker 2: I find it very interesting, so I can possibly couple 421 00:23:19,320 --> 00:23:21,800 Speaker 2: it with this and then I can, you know, find 422 00:23:21,800 --> 00:23:22,360 Speaker 2: some solution. 423 00:23:23,880 --> 00:23:28,520 Speaker 1: Tina's recognition of the unsustainability of facial recognition is very valuable. 424 00:23:29,119 --> 00:23:31,240 Speaker 1: My mother always said the eyes are the windows of 425 00:23:31,280 --> 00:23:34,680 Speaker 1: the soul. But Tina understands that who we are has 426 00:23:34,720 --> 00:23:37,640 Speaker 1: a lot more nuance to it. This is so important 427 00:23:37,680 --> 00:23:41,520 Speaker 1: to how machine learning develops to become more inclusive. One 428 00:23:41,520 --> 00:23:43,960 Speaker 1: of the biggest concerns with AI is a distrust of 429 00:23:43,960 --> 00:23:47,680 Speaker 1: the machine's ability to understand humanity. What is great about 430 00:23:47,680 --> 00:23:50,080 Speaker 1: hearing Tina speak is that her work is rooted in 431 00:23:50,160 --> 00:23:53,919 Speaker 1: finding multiple ways to understand humans. This gives me a 432 00:23:53,920 --> 00:23:56,000 Speaker 1: lot of hope for what AI can be, and we 433 00:23:56,119 --> 00:24:02,040 Speaker 1: have people like Tina behind its development. Just to circle 434 00:24:02,080 --> 00:24:04,680 Speaker 1: back round at the start, we talked about the start 435 00:24:04,720 --> 00:24:08,200 Speaker 1: of your story and getting inspired by the AI Youth 436 00:24:08,400 --> 00:24:11,920 Speaker 1: program run by Intel. I'd like to get a sense 437 00:24:11,960 --> 00:24:15,880 Speaker 1: of in terms of your peer group, how much interest 438 00:24:16,160 --> 00:24:20,280 Speaker 1: is there in AI development and STEM. I guess in 439 00:24:20,359 --> 00:24:23,680 Speaker 1: your coh of friends and peers, is it something they're 440 00:24:23,720 --> 00:24:27,240 Speaker 1: interested in and do you see a trend growing or 441 00:24:27,240 --> 00:24:29,440 Speaker 1: are there's still more challenges for people to take up 442 00:24:29,560 --> 00:24:32,080 Speaker 1: that sort of role in their career. 443 00:24:32,760 --> 00:24:35,399 Speaker 2: Whatever peer groups I have, they all of them are 444 00:24:35,480 --> 00:24:38,879 Speaker 2: quite interested in data science and machine learning. We know 445 00:24:38,920 --> 00:24:41,600 Speaker 2: the data is the new oil, so like there are 446 00:24:41,600 --> 00:24:44,240 Speaker 2: a huge number of job rules that have been coming up. 447 00:24:44,520 --> 00:24:47,080 Speaker 2: And since many of my friends and acquaintances we are 448 00:24:47,119 --> 00:24:50,639 Speaker 2: like financially weak, so all of us look towards earning 449 00:24:50,680 --> 00:24:55,000 Speaker 2: some skill set and becoming job ready, increasing unemployability rather 450 00:24:55,040 --> 00:24:57,240 Speaker 2: than you know, we do not focus on taking this 451 00:24:57,480 --> 00:24:59,840 Speaker 2: up on a longer run. So, I mean there's a 452 00:24:59,880 --> 00:25:03,280 Speaker 2: lot up in this because we know that machine learning, 453 00:25:03,359 --> 00:25:06,440 Speaker 2: artificial intelligence, deep learning and whatever technologies that they are coming up, 454 00:25:06,680 --> 00:25:09,880 Speaker 2: they hold the potential to change, to transform the landscape 455 00:25:09,880 --> 00:25:12,560 Speaker 2: of every industry. So if we take it up as 456 00:25:12,560 --> 00:25:15,199 Speaker 2: a profession, then we need to stay in it for 457 00:25:15,240 --> 00:25:17,959 Speaker 2: a long run. But there are a multitude of impairments 458 00:25:18,000 --> 00:25:19,800 Speaker 2: to it. So the very first one is like I 459 00:25:19,960 --> 00:25:23,359 Speaker 2: being a girl. So in India, like especially from the 460 00:25:23,400 --> 00:25:26,240 Speaker 2: place I belong to, girls are usually not encouraged to 461 00:25:26,240 --> 00:25:29,239 Speaker 2: take up STEM fields. So we need to overcome that 462 00:25:29,320 --> 00:25:32,560 Speaker 2: first of all. And then once we become employable, once 463 00:25:32,560 --> 00:25:36,080 Speaker 2: we become like financially stable independent, I mean, then talking 464 00:25:36,119 --> 00:25:38,440 Speaker 2: on a personal level, I can then you know, work 465 00:25:38,560 --> 00:25:41,080 Speaker 2: in this field, and then I can possibly work in 466 00:25:41,119 --> 00:25:45,040 Speaker 2: somewhere around mental health and machine learning. And therefore, in 467 00:25:45,080 --> 00:25:48,480 Speaker 2: the coming future I plan to you know, launch a 468 00:25:48,520 --> 00:25:52,240 Speaker 2: program to say which is shakti in STEM. So Shakti 469 00:25:52,400 --> 00:25:54,399 Speaker 2: is a Hindi word and a literal meaning. It means 470 00:25:54,400 --> 00:25:57,760 Speaker 2: feminine energy. Apart from this, it also has a different meaning, 471 00:25:57,800 --> 00:26:00,760 Speaker 2: like in India, Shakti is used to represent strong and 472 00:26:00,800 --> 00:26:03,840 Speaker 2: resilient young girls and women. So I would want to 473 00:26:03,920 --> 00:26:07,800 Speaker 2: launch this program Shuck teen Stem, which aims at educating 474 00:26:07,880 --> 00:26:11,840 Speaker 2: youngers who are based in rural areas who heal from 475 00:26:11,880 --> 00:26:15,960 Speaker 2: financially weaker and economically weaker and socially backward start of 476 00:26:15,960 --> 00:26:18,880 Speaker 2: the society and to educate them and to fuel their 477 00:26:18,880 --> 00:26:21,320 Speaker 2: aspirations to enter into STEM careers. 478 00:26:22,040 --> 00:26:24,880 Speaker 1: Yeah, that's awesome because I mean, I have two daughters 479 00:26:24,920 --> 00:26:28,320 Speaker 1: and I'm really encouraging them to get into the STEM 480 00:26:28,600 --> 00:26:31,840 Speaker 1: side of things. And you know, anything to help anyone 481 00:26:31,880 --> 00:26:36,400 Speaker 1: get into coding and developing and actually creating something from 482 00:26:36,400 --> 00:26:39,960 Speaker 1: new is quite a exciting feeling. So thanks Tina for 483 00:26:40,040 --> 00:26:43,120 Speaker 1: joining us today. I really enjoyed that and I learned 484 00:26:43,160 --> 00:26:43,919 Speaker 1: quite a lot from this. 485 00:26:44,280 --> 00:26:45,520 Speaker 2: Thank you, Thank you. 486 00:26:50,680 --> 00:26:53,399 Speaker 1: Thank you to my guest Tina Sahu for joining me 487 00:26:53,560 --> 00:26:58,360 Speaker 1: on this episode of Technically Speaking, an Intel podcast. This 488 00:26:58,480 --> 00:27:01,720 Speaker 1: episode brilliantly highlighted the potential of AI in supporting those 489 00:27:01,760 --> 00:27:05,080 Speaker 1: facing mental health challenges. I firmly believe that within the 490 00:27:05,119 --> 00:27:08,680 Speaker 1: next decade will witness a surge in AI powered therapeutic 491 00:27:08,720 --> 00:27:12,760 Speaker 1: tools designed especially for the younger generation navigating life hurdles. 492 00:27:13,600 --> 00:27:17,280 Speaker 1: One heartening development is society's evolving recognition of mental health 493 00:27:17,480 --> 00:27:20,679 Speaker 1: as a genuine concern. I remember the nineties as a 494 00:27:20,720 --> 00:27:23,840 Speaker 1: fresh faced teenager. It was a time when such discussions 495 00:27:23,880 --> 00:27:27,159 Speaker 1: were almost taboo and laden with stigma. Yet there's a 496 00:27:27,200 --> 00:27:30,800 Speaker 1: pressing issue the shortage of well trained mental health professionals 497 00:27:31,240 --> 00:27:35,040 Speaker 1: to cater to the increasing demand. AI and tech can 498 00:27:35,080 --> 00:27:39,080 Speaker 1: serve as invaluable aids for these professionals, ultimately benefiting our 499 00:27:39,119 --> 00:27:43,679 Speaker 1: community at large. Tina's transition from technology skeptic to its 500 00:27:43,800 --> 00:27:46,600 Speaker 1: ardent supporter was a highlight for me as a father 501 00:27:46,680 --> 00:27:49,359 Speaker 1: of three. I'm hopeful not just about the job prospects 502 00:27:49,400 --> 00:27:52,359 Speaker 1: AI will offer them, but also the tech savvy liars 503 00:27:52,359 --> 00:27:55,600 Speaker 1: they will lead, with AI becoming second nature to them. 504 00:27:56,080 --> 00:27:59,680 Speaker 1: Observing the innovative solutions emerging from young minds like Tina's, 505 00:28:00,160 --> 00:28:04,120 Speaker 1: I'm convinced we're on the cusps discovering awesome new technologies, apps, 506 00:28:04,160 --> 00:28:09,680 Speaker 1: and remedies for many of life's challenges. Please join us 507 00:28:09,720 --> 00:28:13,400 Speaker 1: on Tuesday, November twenty eighth for the next two episodes 508 00:28:13,480 --> 00:28:17,400 Speaker 1: of Technically Speaking, an Intel podcast we'll be sharing two 509 00:28:17,400 --> 00:28:21,720 Speaker 1: special episodes exploring the future of transportation and how technology 510 00:28:21,920 --> 00:28:25,680 Speaker 1: like AI has already created modern day and mobility marvels 511 00:28:26,200 --> 00:28:35,199 Speaker 1: like flying cars and autonomous shuttles. Technically Speaking, was produced 512 00:28:35,200 --> 00:28:38,640 Speaker 1: by Ruby Studios from iHeartRadio in partnership with Intel, and 513 00:28:38,680 --> 00:28:43,000 Speaker 1: hosted by me Graham Class. Our executive producer is Moley Sosha, 514 00:28:43,400 --> 00:28:46,160 Speaker 1: our ep of Post production is James Foster, and our 515 00:28:46,200 --> 00:28:50,640 Speaker 1: supervising producer is Nikias Swinton. This episode was edited by 516 00:28:50,680 --> 00:29:02,880 Speaker 1: Cira Spreen and written and produced by Tiree Rush.