WEBVTT 00:00:00.000 --> 00:00:02.180 We're going to talk about artificial intelligence 00:00:02.180 --> 00:00:06.500 and we have Carlos Colonga from where again? 00:00:06.780 --> 00:00:09.580 Hospital Regional Materno Infantil de Especialidades 00:00:09.580 --> 00:00:16.339 de Monterrey. He's from Mexico. We have Em Gootee, 00:00:16.579 --> 00:00:20.440 who came as my medical media fellow from Turkey 00:00:20.440 --> 00:00:23.519 after graduating medical school, and I was fortunate 00:00:23.519 --> 00:00:26.640 to recruit her to stay on as one of my colleagues 00:00:26.640 --> 00:00:28.960 now at Cincinnati Children's, helping us in innovation. 00:00:29.339 --> 00:00:32.159 And Rami Shahaban, who I was introduced to years 00:00:32.159 --> 00:00:36.600 ago, who is an ENT surgeon who is now a Ph .D. 00:00:36.600 --> 00:00:40.039 and a professor in instructional technology at 00:00:40.039 --> 00:00:43.420 Utah State University. Rock on. All right. This 00:00:43.420 --> 00:00:45.380 is the team. They're going to update us on artificial 00:00:45.380 --> 00:01:02.619 intelligence. Take it away. So we are going to 00:01:02.619 --> 00:01:07.040 talk about AI, specifically regenerative AI use 00:01:07.040 --> 00:01:10.859 cases in medicine and medical education. and 00:01:10.859 --> 00:01:14.540 we have three categories of use cases research 00:01:14.540 --> 00:01:19.469 productivity and medical education So let's start 00:01:19.469 --> 00:01:23.450 with that poll question here. So which of the 00:01:23.450 --> 00:01:27.189 following is false? Is false. Okay. So that's 00:01:27.189 --> 00:01:30.689 tricky a little bit. ChatGBT can help to organize 00:01:30.689 --> 00:01:34.370 literature review. ChatGBT can analyze data within 00:01:34.370 --> 00:01:38.709 an Excel sheet. ChatGBT can run statistical analyses. 00:01:39.049 --> 00:01:42.329 ChatGBT can help in the research design. Or none 00:01:42.329 --> 00:01:48.359 of the above. So what do you think? Any thoughts 00:01:48.359 --> 00:01:53.579 here in the room? Do you think any of these are 00:01:53.579 --> 00:01:56.760 false or all of them are true? What about the 00:01:56.760 --> 00:02:01.459 unpaid version? It still doesn't. Yeah. Mike. 00:02:02.760 --> 00:02:07.959 Yeah. Oh, right. Yeah. We're talking about both 00:02:07.959 --> 00:02:13.090 of them, yes. Let's keep going. All of them are 00:02:13.090 --> 00:02:18.469 true. We are researching how ChatGPT can help 00:02:18.469 --> 00:02:22.430 us in the research process. I personally use 00:02:22.430 --> 00:02:25.650 ChatGPT in systematic review and meta -analysis. 00:02:27.490 --> 00:02:33.370 When you do systematic review, you screen thousands 00:02:33.370 --> 00:02:37.389 of articles, and then you have a title and abstract 00:02:37.389 --> 00:02:40.849 screening, and then the full text screen. And 00:02:40.849 --> 00:02:45.669 this process is intimidating. You have to have 00:02:45.669 --> 00:02:50.569 many people involved in that process. So we tried 00:02:50.569 --> 00:02:53.229 ChatGPT to do the title and abstract screening, 00:02:53.490 --> 00:02:56.090 as well as the full text screening, and it does 00:02:56.090 --> 00:02:58.629 a great job. Do you want to explain how to do 00:02:58.629 --> 00:03:03.449 it? Yeah, sure. I think we have the next slide 00:03:03.449 --> 00:03:07.680 here. Talking about that so we have punch of 00:03:07.680 --> 00:03:10.520 articles to go into title abstract screen And 00:03:10.520 --> 00:03:14.840 then we asked chat GVT this prompt. I'm conducting 00:03:14.840 --> 00:03:17.120 a systematic review of the use of AI in organ 00:03:17.120 --> 00:03:21.800 transplantation I need you to act as a data analyst 00:03:21.800 --> 00:03:24.680 Please analyze the attached excel sheet and screen 00:03:24.680 --> 00:03:27.319 the included articles for title abstract by the 00:03:27.319 --> 00:03:30.939 way that The AI in organ transplantation paper 00:03:30.939 --> 00:03:34.550 was published already so we used chatgbt in that 00:03:34.550 --> 00:03:37.110 use the following inclusion criteria we have 00:03:37.110 --> 00:03:41.229 some inclusion criteria here and then after screening 00:03:41.229 --> 00:03:45.009 the title and abstract so now we want chatgbt 00:03:45.009 --> 00:03:48.110 also to manipulate the document itself so it's 00:03:48.110 --> 00:03:52.430 not only just giving me text in a answer through 00:03:52.430 --> 00:03:56.590 chatgbt but also go inside the excel sheet and 00:03:56.590 --> 00:04:00.770 then read the data from excel sheet add columns 00:04:00.770 --> 00:04:04.780 to the Excel sheet and say whether this article 00:04:04.780 --> 00:04:07.500 is relevant or not relevant or maybe relevant. 00:04:07.800 --> 00:04:11.639 Because we don't trust ChatGPT a lot, I asked 00:04:11.639 --> 00:04:16.139 ChatGPT also to add a column to explain why ChatGPT 00:04:16.139 --> 00:04:20.360 decided that decision. So you have a better understanding 00:04:20.360 --> 00:04:24.139 about this data and if you don't agree with ChatGPT, 00:04:24.220 --> 00:04:26.379 you can manipulate the prompt, engineer the prompt 00:04:26.379 --> 00:04:29.540 to make it more accurate. Okay, and we get an 00:04:29.540 --> 00:04:33.579 excellent result with that. We had to do the 00:04:33.579 --> 00:04:36.920 human title and abstract screening also because 00:04:36.920 --> 00:04:41.800 we want to have a reliable answer, but we did 00:04:41.800 --> 00:04:45.540 inter -rater reliability between ChagVT and the 00:04:45.540 --> 00:04:50.740 human reliability, and the inter -rater reliability 00:04:50.740 --> 00:04:53.839 was excellent. So ChagVT could be used in that. 00:04:54.839 --> 00:04:57.079 part yeah so what is what is the first cycle 00:04:57.079 --> 00:05:02.319 and second so yeah okay so uh first we wanted 00:05:02.319 --> 00:05:06.240 to use chat gbt uh to give us yes or no question 00:05:06.240 --> 00:05:09.439 yes or no relevant or not relevant and then we 00:05:09.439 --> 00:05:12.899 just redesigned the prompt a little bit to make 00:05:12.899 --> 00:05:15.620 chat gbt say whether it's relevant not relevant 00:05:15.620 --> 00:05:18.920 and maybe relevant so we decided to consider 00:05:18.920 --> 00:05:22.399 all of these versions of the prompt and then 00:05:22.399 --> 00:05:24.550 make the inter -rater reliability between all 00:05:24.550 --> 00:05:30.149 of that columns. Question? I've done this, not 00:05:30.149 --> 00:05:33.029 for the exact same purpose, but I don't get the 00:05:33.029 --> 00:05:36.550 same answer each time. So my question is about 00:05:36.550 --> 00:05:39.310 reliability. I know humans are not reliable either, 00:05:39.449 --> 00:05:42.829 but it sounds like every time you offer this 00:05:42.829 --> 00:05:44.870 question, at least to ChatGPT, I haven't tried 00:05:44.870 --> 00:05:48.750 other ones, you get a slightly different answer. 00:05:48.889 --> 00:05:51.100 It approaches the problem like... for the first 00:05:51.100 --> 00:05:53.759 time every time? I think so. Yeah, so if you 00:05:53.759 --> 00:05:56.740 use ChatGPT in the general search, you get different 00:05:56.740 --> 00:05:59.899 answers, but we created custom GPT for that. 00:06:00.019 --> 00:06:04.680 So you can walk ChatGPT in serial of prompts 00:06:04.680 --> 00:06:08.319 until you get the most accurate results, which 00:06:08.319 --> 00:06:11.899 is just not that difference between if you run 00:06:11.899 --> 00:06:14.439 it multiple times, you get a slight difference. 00:06:14.660 --> 00:06:18.519 And we had to have human reliability here also 00:06:18.519 --> 00:06:23.050 to just make sure. that it is reliable. But that's 00:06:23.050 --> 00:06:26.550 a good potential for the future iterations for 00:06:26.550 --> 00:06:30.649 . I think also it might be one of the questions 00:06:30.649 --> 00:06:33.509 that, what is the AI limitations when you use 00:06:33.509 --> 00:06:36.459 it for an actual paper? and go for publishing, 00:06:36.699 --> 00:06:39.959 what is the disclosures that you do for the journals, 00:06:40.199 --> 00:06:41.800 do they have any limitations, do you have any 00:06:41.800 --> 00:06:45.199 extra information on that? Yeah, so we use ChatGPT 00:06:45.199 --> 00:06:48.100 here as a part of the methodology and part of 00:06:48.100 --> 00:06:51.120 the reliability process. We don't use it to, 00:06:51.300 --> 00:06:55.779 for example, as an act of plagiarism or writing, 00:06:56.000 --> 00:07:00.139 yes. So, yeah, our journals are very, yeah, of 00:07:00.139 --> 00:07:04.519 course, yeah. And we frankly say that this cycle 00:07:04.810 --> 00:07:07.870 of screening is done by . The second cycle of 00:07:07.870 --> 00:07:10.209 screening done by . Third cycle of screening 00:07:10.209 --> 00:07:12.970 is done by humans. So, yeah. You always have 00:07:12.970 --> 00:07:16.110 to have a human in the equations. So we have 00:07:16.110 --> 00:07:20.389 to take in mind that these AI models, they help 00:07:20.389 --> 00:07:23.050 us increase our efficiency. And sometimes we 00:07:23.050 --> 00:07:26.910 equate wrongly efficiency to productivity, but 00:07:26.910 --> 00:07:29.930 sometimes you get productivity out of the efficiency. 00:07:30.069 --> 00:07:33.420 So it helps us be more efficient. by accelerating 00:07:33.420 --> 00:07:36.399 how we do the work. But we always, at least in 00:07:36.399 --> 00:07:38.899 the medical and the science part, we have to 00:07:38.899 --> 00:07:41.060 have a human in the equation. Because we have 00:07:41.060 --> 00:07:43.800 to remember that even if we use the prompts that 00:07:43.800 --> 00:07:47.759 help you get the same results, because ChatGPT 00:07:47.759 --> 00:07:49.939 is a black box model that we don't know what 00:07:49.939 --> 00:07:52.420 it is trained about, and they're always changing 00:07:52.420 --> 00:07:55.600 the weights, it'll change. So it needs us to 00:07:55.600 --> 00:07:58.920 tell it what to do and to understand and read 00:07:58.920 --> 00:08:02.449 the answers of ChatGPT. Yeah, you want to move 00:08:02.449 --> 00:08:05.829 on to the next one? Yeah. So let's go to talk 00:08:05.829 --> 00:08:09.810 about Jenny AI, whose motto is use AI to supercharge 00:08:09.810 --> 00:08:13.230 your research paper. So what does it do? You 00:08:13.230 --> 00:08:16.370 open the Jenny AI, which is one of different 00:08:16.370 --> 00:08:21.449 academic models that has been fine -tuned with 00:08:21.449 --> 00:08:24.040 research papers. And you literally put your prompt. 00:08:24.199 --> 00:08:26.860 I want to write about this, this, and that. In 00:08:26.860 --> 00:08:28.899 this example, let's talk about pediatric interception. 00:08:29.160 --> 00:08:31.399 You put the prompt. And the first thing that 00:08:31.399 --> 00:08:34.379 you'll get is a suggestion. You can start writing 00:08:34.379 --> 00:08:37.940 this. You, because you're the surgeon that has 00:08:37.940 --> 00:08:41.059 the hand on the wheel, can say, yeah, I like 00:08:41.059 --> 00:08:43.139 it. I want to modify it. I want to change it 00:08:43.139 --> 00:08:47.100 in this and that way. And then, the next slide, 00:08:47.159 --> 00:08:51.259 please. You can also talk with its in -house 00:08:51.259 --> 00:08:54.769 chatbot. and say, hey, I really want to give 00:08:54.769 --> 00:08:58.409 it this pin. I want it to research specifically 00:08:58.409 --> 00:09:02.149 in this. And the chatbot will give you the answers 00:09:02.149 --> 00:09:05.889 that you're searching for. It will cite papers 00:09:05.889 --> 00:09:10.009 to you, and it can help you cite on your preferred 00:09:10.009 --> 00:09:14.690 type of citation specifications into your paper. 00:09:15.330 --> 00:09:19.720 Next, please. So here. You've decided that, yes, 00:09:19.799 --> 00:09:22.700 I like what it is saying. I also want you to 00:09:22.700 --> 00:09:25.860 add this, this, and that on my academic paper. 00:09:26.200 --> 00:09:28.700 And it gives you a little suggestion of the summary 00:09:28.700 --> 00:09:31.200 of the papers that you want to introduce. So 00:09:31.200 --> 00:09:34.399 we can see how this accelerates your paper writing, 00:09:34.519 --> 00:09:36.460 but you're still at the command of it. You're 00:09:36.460 --> 00:09:42.600 still in the helm. Next, please. So let's do 00:09:42.600 --> 00:09:44.720 an audience poll to see how you're feeling about 00:09:44.720 --> 00:09:48.019 this. So how recent are the articles suggested 00:09:48.019 --> 00:09:52.200 from the use of these AI models? So they have 00:09:52.200 --> 00:09:55.480 a cut -off date of 2023. They're older than five 00:09:55.480 --> 00:10:00.500 years, older than 10 years, and they're on par 00:10:00.500 --> 00:10:02.580 with the latest published research. What do you 00:10:02.580 --> 00:10:13.179 think? Any ideas? Can we see the answers from 00:10:13.179 --> 00:10:15.320 the international audience, please? I don't know 00:10:15.320 --> 00:10:18.100 if the poll is up. Oh. Can we see the poll results? 00:10:19.279 --> 00:10:26.220 Okay. How do people know what's up? Yeah. So, 00:10:26.220 --> 00:10:29.600 mostly, nowadays, they're mostly on par. We just 00:10:29.600 --> 00:10:31.519 have to research into what model we're using 00:10:31.519 --> 00:10:33.820 and what library they're using, but it is on 00:10:33.820 --> 00:10:36.080 par. We just have to remember that they're searching. 00:10:36.620 --> 00:10:39.019 for those latest articles in this specific type 00:10:39.019 --> 00:10:42.500 of library, which we'll talk about it at a later 00:10:42.500 --> 00:10:45.159 moment. Well, that was the moment, yeah. So... 00:10:45.159 --> 00:10:48.120 Also, there's another tool with the Gen -E AI. 00:10:48.220 --> 00:10:51.299 It's called Open Evidence, and I think you said 00:10:51.299 --> 00:10:54.179 this is the difference that this is not necessarily 00:10:54.179 --> 00:10:57.340 helping you write it, but it helps you do literature 00:10:57.340 --> 00:11:01.379 search. Exactly, which is amazing. This has been... 00:11:01.740 --> 00:11:04.720 This was part of the program accelerator over 00:11:04.720 --> 00:11:08.440 by Harvard team and the Magic Clinic. And they 00:11:08.440 --> 00:11:12.679 especially focus on the Elsevier Library. So 00:11:12.679 --> 00:11:15.940 you can ask any type of question. It will reply 00:11:15.940 --> 00:11:19.139 in a conversational manner, trying to answer 00:11:19.139 --> 00:11:23.600 your question while summarizing the papers, the 00:11:23.600 --> 00:11:27.700 latest papers that it's citing. So it gives you 00:11:27.700 --> 00:11:30.620 a conversational answer. And then at the end 00:11:30.620 --> 00:11:33.299 of it, It gives you the papers that you can personally 00:11:33.299 --> 00:11:36.940 check out to confirm what it's saying. If you 00:11:36.940 --> 00:11:39.320 want to press the details, it gives you a bigger 00:11:39.320 --> 00:11:43.360 summary. And then it suggests, hey, so you're 00:11:43.360 --> 00:11:44.480 talking about these things. It's like you just 00:11:44.480 --> 00:11:46.580 cited Dr. Russell's paper in the second part. 00:11:47.000 --> 00:11:49.980 Yeah, yeah. And if you wanted to go... Accurate. 00:11:50.200 --> 00:11:53.580 Yeah. And it also gives you suggestions. So I 00:11:53.580 --> 00:11:55.259 know you're talking about the appendectomy. Hey, 00:11:55.340 --> 00:11:56.940 you could also talk about this and that, which 00:11:56.940 --> 00:12:01.220 is up in the next slide. And you can ask follow 00:12:01.220 --> 00:12:04.399 -up questions. Yep, yep, yeah. So it's pretty 00:12:04.399 --> 00:12:10.080 great. So what would you say the personal practice 00:12:10.080 --> 00:12:15.799 changes for your part? For my part? Okay. So 00:12:15.799 --> 00:12:22.399 we are going to talk about the medical education 00:12:22.399 --> 00:12:25.470 part. Is that right? Medical education is coming 00:12:25.470 --> 00:12:28.309 next. Okay. We're going to first talk about the 00:12:28.309 --> 00:12:30.789 productivity part, but I think it kind of summarizes 00:12:30.789 --> 00:12:35.149 that large language models can enhance the efficiency 00:12:35.149 --> 00:12:38.549 and productivity when you do research by analyzing 00:12:38.549 --> 00:12:41.610 the large data sets for you, or you can use those 00:12:41.610 --> 00:12:44.629 tools to help you write. This is especially very 00:12:44.629 --> 00:12:46.710 important for the international audience that 00:12:46.710 --> 00:12:51.970 have, once you have something on hand and work 00:12:51.970 --> 00:12:54.210 on it, it's way easier. and writing something 00:12:54.210 --> 00:12:57.990 by yourself from scratch. So these large language 00:12:57.990 --> 00:13:00.450 models also can be an editor and help you write 00:13:00.450 --> 00:13:04.850 these papers. And we're gonna move on. Yeah, 00:13:04.929 --> 00:13:10.309 Jose. Dr. Holcomb, what's the JPS policy about 00:13:10.309 --> 00:13:18.250 these AI tools? I think we wanted to submit something. 00:13:18.370 --> 00:13:22.659 They asked if we used it. Yeah, so you can use 00:13:22.659 --> 00:13:26.220 it for background information, but you can't 00:13:26.220 --> 00:13:30.000 use generative AI, that is, to have AI write 00:13:30.000 --> 00:13:33.679 you the paper. And are you making sure it's not 00:13:33.679 --> 00:13:35.500 being used, or you have to disclose it somewhere? 00:13:35.720 --> 00:13:39.740 Well, first of all, if you used it and you disclosed 00:13:39.740 --> 00:13:41.460 it, it's probably not going to get accepted. 00:13:42.700 --> 00:13:47.840 But we have, we being Elsevier, has tools to... 00:13:48.000 --> 00:13:51.159 to figure out if it's been used or if it's being 00:13:51.159 --> 00:13:55.820 used. So that's sort of the policy is you can't 00:13:55.820 --> 00:13:59.000 use generative AI to create your paper, to write 00:13:59.000 --> 00:14:03.120 your paper. I think it's okay to use it. You 00:14:03.120 --> 00:14:07.759 could, for language, it's probably appropriate. 00:14:08.320 --> 00:14:11.759 And for literature, you know, background information, 00:14:12.120 --> 00:14:15.250 it's certainly appropriate. I think it all started 00:14:15.250 --> 00:14:17.789 a couple years ago when ChatGPT started getting 00:14:17.789 --> 00:14:21.750 used in different fields. The first couple papers 00:14:21.750 --> 00:14:25.250 were written and ChatGPT was among the authors. 00:14:25.590 --> 00:14:29.570 And I think it's not been permitted anymore. 00:14:29.889 --> 00:14:32.669 You cannot do that. But I don't know. Probably 00:14:32.669 --> 00:14:34.710 it's the same for all the journals. You cannot 00:14:34.710 --> 00:14:37.850 ask ChatGPT to write a section of it and show 00:14:37.850 --> 00:14:40.639 it in the authors. Yeah, most of the journals 00:14:40.639 --> 00:14:44.340 I see is just asking you to, if you want to use 00:14:44.340 --> 00:14:47.299 ChatGPT, you use it to just enhance the writing 00:14:47.299 --> 00:14:53.139 style, not create the ideas. So, yeah. And our 00:14:53.139 --> 00:14:55.519 next question, because we're going to talk about 00:14:55.519 --> 00:14:58.840 ChatGPT and other AI tools and how you use to 00:14:58.840 --> 00:15:01.259 enhance your daily productivity, not necessarily 00:15:01.259 --> 00:15:03.460 for research. And we asked the audience, how 00:15:03.460 --> 00:15:05.500 often do you use ChatGPT or similar AI tools? 00:15:05.600 --> 00:15:09.620 It could be Cloud, Gemini, anything. It's a spectrum 00:15:09.620 --> 00:15:13.100 from never to always. And one third is never. 00:15:13.399 --> 00:15:16.580 Whenever Todd sees those numbers. He kind of 00:15:16.580 --> 00:15:19.799 has a heart attack every time he sees people 00:15:19.799 --> 00:15:33.200 that never use chat, GPT, or AI. So if you first 00:15:33.200 --> 00:15:36.639 start with productivity part with MeetGeek, this 00:15:36.639 --> 00:15:39.700 is one of the tools that helps you. You basically 00:15:39.700 --> 00:15:42.200 synchronize this tool with your calendar. It 00:15:42.200 --> 00:15:46.250 jumps in. Meeting that you attach to your calendar 00:15:46.250 --> 00:15:48.789 and it takes notes for you and at the end of 00:15:48.789 --> 00:15:51.990 it It sends you a mail of the transcription notes 00:15:51.990 --> 00:15:56.830 and the summary And it basically tells like who 00:15:56.830 --> 00:15:59.110 was in the Hoover in the meeting what was the 00:15:59.110 --> 00:16:03.110 summary and what you need to do it goes with 00:16:03.110 --> 00:16:06.649 the recording and it assigns action items per 00:16:06.649 --> 00:16:10.169 person who are in the meaning and Fireflies is 00:16:10.169 --> 00:16:13.779 a similar one with the Miki if you go look in 00:16:13.779 --> 00:16:17.639 the internet to see meeting summarizer AI tools. 00:16:17.899 --> 00:16:20.000 They're like a bunch of them. Most of them does 00:16:20.000 --> 00:16:22.360 the same job. They're on different price points. 00:16:23.240 --> 00:16:27.500 Most of the time, it's just personal preference. 00:16:27.700 --> 00:16:31.220 They mostly do the same thing. I know you started 00:16:31.220 --> 00:16:33.759 using MeetGeek and then you transferred to Fireflies 00:16:33.759 --> 00:16:36.620 because we... Influenced you? Yeah, I think Firefly, 00:16:36.639 --> 00:16:39.580 yeah. So I use mid -gig, I use the paid version 00:16:39.580 --> 00:16:41.899 of mid -gig, and then Firefly is the free version, 00:16:41.980 --> 00:16:44.259 and then the paid version. And when you invite 00:16:44.259 --> 00:16:49.840 Rami to a meeting, shows up as four people. Rami, 00:16:49.899 --> 00:16:51.639 AI note -taker one, Rami, AI note -taker two. 00:16:51.639 --> 00:16:53.500 The human version is not there, is that right? 00:16:54.440 --> 00:16:59.019 So I think Firefly is... My personal preference 00:16:59.019 --> 00:17:03.980 is because it has also action items that So after 00:17:03.980 --> 00:17:06.779 the meeting it it gives you the action items 00:17:06.779 --> 00:17:10.680 for what's next so it's just to keep you Informed 00:17:10.680 --> 00:17:12.720 it is great. I use fireflies because of you guys. 00:17:12.759 --> 00:17:16.880 I started using Carlos also comes as a two people 00:17:16.880 --> 00:17:20.890 every meeting The issue that we've run into, 00:17:21.049 --> 00:17:23.150 because I was addicted to it, I needed it because 00:17:23.150 --> 00:17:26.150 it gives me takeaways and all the key points. 00:17:26.250 --> 00:17:28.990 Here's what you need to do. But then our hospital, 00:17:29.329 --> 00:17:32.430 people got worried, especially with some of the 00:17:32.430 --> 00:17:34.730 meetings I was having, that it was auto -recording 00:17:34.730 --> 00:17:38.109 everything. So you just have to figure out what's 00:17:38.109 --> 00:17:40.549 the safe way. At Cincinnati Children's, they 00:17:40.549 --> 00:17:44.559 now have... co -pilot approved by co -pilot that's 00:17:44.559 --> 00:17:46.920 approved by the hospital but even then our lawyers 00:17:46.920 --> 00:17:50.259 asked me to stop using it on half of my meetings 00:17:50.259 --> 00:17:52.400 because they're legally sensitive well you talk 00:17:52.400 --> 00:17:54.700 about confidential things yeah confidential things 00:17:54.700 --> 00:17:58.569 so um it's it's We're going to make progress 00:17:58.569 --> 00:18:00.869 with that because this is such a valuable tool, 00:18:00.950 --> 00:18:03.490 and we have to get past the security part of 00:18:03.490 --> 00:18:07.130 it, but it is game -changing. Especially if you're 00:18:07.130 --> 00:18:09.390 using an institution level, it needs to be a 00:18:09.390 --> 00:18:13.730 lot of preventative, like preventions be taking 00:18:13.730 --> 00:18:16.430 place for security purposes, especially like 00:18:16.430 --> 00:18:19.049 you and Dr. Winallman's level executive meetings, 00:18:19.269 --> 00:18:23.089 but like you and me in a meeting. Fireflies can 00:18:23.089 --> 00:18:25.430 record that. There's not going to be that one. 00:18:26.750 --> 00:18:30.460 We've actually. Added it to our platform. So 00:18:30.460 --> 00:18:32.980 we use a co -pilot and that's the only thing 00:18:32.980 --> 00:18:35.779 that's basically approved You know for Carl stores 00:18:35.779 --> 00:18:39.359 to use in regards to note -taking or getting 00:18:39.359 --> 00:18:42.759 information Recording stuff and then you know 00:18:42.759 --> 00:18:44.839 analyzing it afterwards I think it's the same 00:18:44.839 --> 00:18:47.559 at Cincinnati Children's and they keeps that 00:18:47.559 --> 00:18:50.940 they Generally send us like weekly AI guidelines 00:18:50.940 --> 00:18:54.339 of how should we use it? And what kind of information 00:18:54.339 --> 00:18:57.500 should not be used when you use co -pilot or 00:18:57.500 --> 00:18:59.410 ask for it? meeting summaries and everything. 00:18:59.670 --> 00:19:05.029 And I think that's a good step to endorse people 00:19:05.029 --> 00:19:09.250 to use with not giving away the secrets of the 00:19:09.250 --> 00:19:11.970 company, basically. Yeah, we have to have these 00:19:11.970 --> 00:19:14.109 talks. We have to learn how to use it because 00:19:14.109 --> 00:19:18.150 it's here to stay. We can't close our eyes and 00:19:18.150 --> 00:19:21.369 don't admit that everyone's using it to increase 00:19:21.369 --> 00:19:24.630 their efficiency and improve the way they manage 00:19:24.630 --> 00:19:28.180 their day -to -day lives. Zoom AI basically does 00:19:28.180 --> 00:19:30.579 the same thing because some companies have Zoom 00:19:30.579 --> 00:19:33.880 instead of Microsoft. But the plus I saw in the 00:19:33.880 --> 00:19:37.420 Zoom, if you're late to a meeting and you have 00:19:37.420 --> 00:19:39.960 Zoom AI companion, you can ask if your name is 00:19:39.960 --> 00:19:42.779 mentioned, what people ask you to do, and recap 00:19:42.779 --> 00:19:47.519 the meeting just before you join what happened. 00:19:47.740 --> 00:19:50.740 So I think that's a very cool thing. And if you're 00:19:50.740 --> 00:19:52.799 generally late to the meetings, I would definitely 00:19:52.799 --> 00:20:02.140 recommend. One minute? Okay, sorry. Okay, this 00:20:02.140 --> 00:20:04.859 is Todd's favorite. Read aloud GPT. We created 00:20:04.859 --> 00:20:07.759 for it. You can use this QR code and you can 00:20:07.759 --> 00:20:11.640 use it by yourself. It basically helps you listen 00:20:11.640 --> 00:20:13.880 to your long emails or any text that you need 00:20:13.880 --> 00:20:16.839 to listen to instead of you read. We know Todd 00:20:16.839 --> 00:20:20.809 talks, like have to not talk, sorry. He has to 00:20:20.809 --> 00:20:23.250 read a lot of things, the emails and everything. 00:20:23.509 --> 00:20:26.329 And sometimes, what do you call walk and talk? 00:20:26.670 --> 00:20:31.609 T 'walk. T 'walk. So it's hard to read when you 00:20:31.609 --> 00:20:34.569 walk. So you listen to it and you use the ChatGPT 00:20:34.569 --> 00:20:36.470 to have conversations. Instead of an audio book, 00:20:37.029 --> 00:20:39.009 I will drive. I'm about to drive to Cincinnati 00:20:39.009 --> 00:20:41.829 after this. And I will talk to ChatGPT for an 00:20:41.829 --> 00:20:44.579 hour about a book. So I'll say, summarize this 00:20:44.579 --> 00:20:46.660 book. And I will go back like a book club. I'll 00:20:46.660 --> 00:20:48.599 say, tell me now, what does it say about this? 00:20:48.640 --> 00:20:50.880 Tell me, let's go into chapter two. I want to 00:20:50.880 --> 00:20:53.180 know this. I don't get this. So I will have a 00:20:53.180 --> 00:20:56.660 conversation while I work out, while I am driving. 00:20:57.119 --> 00:20:59.819 And I think this is one of the most important 00:20:59.819 --> 00:21:04.799 ones for this audience here. Todd said he's getting 00:21:04.799 --> 00:21:06.299 a lot of requests for letter of recommendations. 00:21:06.940 --> 00:21:10.140 And we created this GPT for him. And you can 00:21:10.140 --> 00:21:12.279 use the QR code if you can't put the slides up 00:21:12.279 --> 00:21:16.390 here. But they keep saying we created GPTs. It's 00:21:16.390 --> 00:21:18.289 very easy to do. So if you're using the same 00:21:18.289 --> 00:21:20.349 prompts over and over and it's a series of prompts, 00:21:20.589 --> 00:21:23.349 you can create it and just keep reusing it. So 00:21:23.349 --> 00:21:25.910 when I have a letter that I have to write, it 00:21:25.910 --> 00:21:28.569 goes through M -Media for me. So when you use 00:21:28.569 --> 00:21:30.390 this QR code, you're going to go to this letter 00:21:30.390 --> 00:21:33.309 of recommendation GPT. All you need to do is 00:21:33.309 --> 00:21:36.569 upload the CV cover letter. Why are you writing 00:21:36.569 --> 00:21:37.910 this letter of recommendation? It's going to 00:21:37.910 --> 00:21:40.670 give you a draft. You can start working on it 00:21:40.670 --> 00:21:43.829 and personalize it for the candidate. We do not 00:21:43.829 --> 00:21:46.849 recommend the first thing you get from Chachapit 00:21:46.849 --> 00:21:49.369 to use it, but it's a great start for... It will 00:21:49.369 --> 00:21:54.750 sometimes be over... This person is from heaven. 00:21:54.990 --> 00:22:00.049 I actually almost always come back and say, I 00:22:00.049 --> 00:22:03.470 really don't know this person. I'm giving a reference, 00:22:03.690 --> 00:22:06.089 so please tone it down. And then it's like, this 00:22:06.089 --> 00:22:08.130 person is very good, blah, blah, blah. So you 00:22:08.130 --> 00:22:11.599 have to just tell it what you want. Like with 00:22:11.599 --> 00:22:14.079 Halpern, I'm going to say, please really tone 00:22:14.079 --> 00:22:17.779 this down. He's not that good. Do you want to 00:22:17.779 --> 00:22:21.119 do the medical education real fast? Yeah. Sure. 00:22:21.559 --> 00:22:25.730 Just one minute. We'll just talk quickly about 00:22:25.730 --> 00:22:28.470 that. The medical education part, if you want 00:22:28.470 --> 00:22:32.329 to have a personal trainer, we use also ChatGPT 00:22:32.329 --> 00:22:35.769 and other generative AI in that. You can, for 00:22:35.769 --> 00:22:39.329 example, ask ChatGPT to upload a guideline to 00:22:39.329 --> 00:22:42.049 ChatGPT and then ask ChatGPT to create a training 00:22:42.049 --> 00:22:44.789 session, an interactive training session between 00:22:44.789 --> 00:22:48.069 you and ChatGPT. Can we put the slides up? Thank 00:22:48.069 --> 00:22:53.789 you. Yeah, so you can ask ChatGPT to create a 00:22:53.789 --> 00:22:56.309 training session out of a guideline so it doesn't 00:22:56.309 --> 00:22:59.930 have to go into hallucinations or fabricating 00:22:59.930 --> 00:23:02.869 information. It takes the guideline and then 00:23:02.869 --> 00:23:04.750 it creates a series of interactive questions 00:23:04.750 --> 00:23:08.170 where you can interact with ChatGPT and get trained 00:23:08.170 --> 00:23:10.930 on specific things. So, for example, that's an 00:23:10.930 --> 00:23:14.970 appendicitis case that we ask ChatGPT to create 00:23:14.970 --> 00:23:17.710 problem -solving activities. And also we ask 00:23:17.710 --> 00:23:20.789 ChatGPT... to score these activities and give 00:23:20.789 --> 00:23:22.990 us feedback at the end. So that's the feedback 00:23:22.990 --> 00:23:27.109 that it gets at the end. On a similar note, you 00:23:27.109 --> 00:23:30.730 could also make it do any type of case, a step 00:23:30.730 --> 00:23:32.789 one, step two type of case, any type of case. 00:23:33.089 --> 00:23:36.170 You can ask for it. You do the correct prompt, 00:23:36.390 --> 00:23:39.670 and with that, it'll help you do the medical 00:23:39.670 --> 00:23:41.869 case in a step -by -step manner, and at the end, 00:23:41.890 --> 00:23:44.440 it will give you feedback. All right, this was 00:23:44.440 --> 00:23:46.519 awesome. I let us go way over. That's my fault. 00:23:46.579 --> 00:23:48.740 I'm sorry. AI is cool. Keep using it. This is 00:23:48.740 --> 00:23:51.720 phenomenal. These guys are great. We're going 00:23:51.720 --> 00:23:54.059 to keep doing series of courses throughout the 00:23:54.059 --> 00:23:57.160 year for these kind of questions. But guys, thank 00:23:57.160 --> 00:23:58.240 you so much. That was awesome.