ChatGPT 101: How To Use AI In Your Daily Life
Space: Live Event Content
Published: 2024-07-19
3
Views
0
Likes
0
Shares
0
Comments
Timestops
0:00
Introduction to Chat GPT
Todd Ponsky introduces the session on Chat GPT 101, discussing his background as a pediatric surgeon and the importance of digital media in medicine. He emphasizes the need for basic understanding of Chat GPT for those unfamiliar with it.
12:03
Understanding Generative AI
The presenters explain generative AI, its capabilities, and how it differs from traditional AI. They discuss the evolution of AI and its applications in creating content.
24:07
AI in Medical Education
Discussion on how AI can be utilized to create educational materials and enhance learning experiences in the medical field, including examples of AI-generated videos.
36:11
Poll and Audience Engagement
The presenters engage the audience with a poll about their usage of Chat GPT and other AI tools, highlighting the diverse backgrounds of participants.
48:15
Chat GPT vs. Traditional Search Engines
Comparison between Chat GPT and traditional search engines, focusing on the conversational nature of AI and its ability to provide context-aware responses.
1:00:19
Practical Applications of Chat GPT
Exploration of practical uses for Chat GPT in daily tasks, including email summarization, letter writing, and tutoring, emphasizing its role in increasing productivity.
1:12:23
Future of AI in Medicine
Discussion on the future potential of AI in medicine, including ongoing advancements and the importance of staying updated with technological changes.
1:24:26
Interactive Q&A Session
Open floor for audience questions and suggestions on using Chat GPT, fostering a collaborative learning environment.
Topic overview
ChatGPT 101: How to Use Al in Your Daily Life Online Interactive Workshop Talking Points: • Understanding the capabilities and limitations of ChatGPT • Practicing using ChatGPT for various productivity tasks • Q & A
Intended audience: Healthcare professionals and clinicians.
Categories
Specialty
Disease/Condition
Anatomy/Organ System
Care Context
Population
Topic Format
Keywords
generative AI
chat GPT
medical education
digital media
AI technology
task automation
productivity tools
language processing
educational material
conversational interaction
AI applications
personalized learning
tutoring
medical terminology
innovation in medicine
AI-generated content
research fellow
digital communication
user engagement
AI ethics
Hashtags
#GenerativeAI
#ChatGPT
#MedicalEducation
#DigitalMedia
#AITechnology
#TaskAutomation
#ProductivityTools
#LanguageProcessing
#EducationalMaterial
#ConversationalAI
#AIApplications
#PersonalizedLearning
#Tutoring
#MedicalTerminology
#InnovationInMedicine
#AIGeneratedContent
#ResearchFellow
#DigitalCommunication
#UserEngagement
#AIEthics
Transcript
I love to. Good morning. Good afternoon. Good evening. I am Todd Ponsky, and I'm, I'm Gie, and we are gonna be doing chat GPT 101 today. Um, we're super excited to have a new topic other than just medicine all the time. Um, let me, uh, explain who I am and who Em is. So. I am a pediatric surgeon at Cincinnati Children's and the chief innovation officer and I started as a research fellow. Right now I moved to the innovation area with you and we're figuring out my new position. So I asked them what her title is. She's like, I don't know. So M, um, M is absolute superstar. I met M through, uh, it was an app called Clubhouse. I don't know if anyone remembers that. It's like, uh, audio, uh pandemic. It was a pandemic app, audio social media app. We did a, a session on, uh, digital media and medicine. M is an MD from Turkey who said, Hey, I'll come and learn medical media. She came in my lab at Cincinnati Children's. And absolute unbelievable superstar, teaches me everything, I teach her nothing. And uh she has learned um really everything and now has, we've hired her on to be our medical media, um, education, uh, all, all things specialist. So, um, we're gonna do this event together. And, uh, let me explain why we're doing this and what it is, because I will warn you that If you're like a chat GPT like know-how person, this is not for you. This is like the basics, and the reason we're doing this is because um I see that my mom is in the audience and, and uh she's my, uh, the most important person in the audience. She keeps asking me to spend time with her to teach her how to use chat GPT and she doesn't get why she needs that more than Google. And then everyone else started calling me with the same thing. Lots of my friends, they see how much we use it. I literally, I think you said I'm like super you're, you're getting the most, uh, most out of the chat. I'm getting my money's worth. I use it all day long. So everyone said, Can you show us like how you're using this so much? What do you use it for? Why do I need this more than Google? So we're gonna talk about how, um, we use chat GBT that takes our life 10x more productive, um. Literally, I don't think I could do my job without it now. Um, and because I learned so effectively by talking to chat GPT instead of like Googling an article and having conversations, um, Em has been taking some classes and a lot of this stuff, and so she's gonna start off, um, by teaching and then, The main thing is, please put your comments or questions in the chat, and we'll, we have slides, but we'll go off script and be able to address any questions someone has. But more important suggestions, like how are you using it, because I think the more people, the more people hear like how others use it, they go, oh, that's such a great idea. And, and if you have a, um, idea of how to use chat GPT and you don't know how. Uh, just put it in the comments and maybe we can figure it out and share it in the next event. Perfect. Like, I, I'm, I'm gonna be bold enough to say almost anything could be solved with chat GBT. It is I'm so impressive how, how much you use. I know I'm in love with it. Your chat history goes back like 15 chats a day. I love it. I love it. I do. All right, um, um, let's, let's, uh, so hey, Isa, can we go ahead and share the slides. And um, yeah, we'll start with the poll question. OK. Here we go. So we wanna see how often do you use chat GPT right now. It doesn't have to be just chat GPT. It could be Cloud. It could be Gemini. It could be Bing from Microsoft. Um, just be honest, we're just trying to see, um, we just, we're just trying to see what our audience, uh, using right now, and, uh, we'll try to make you feel a little bit more comfortable using it more after this workshop. All right, so I'm looking at the polls. Oh there. I love it, Carolyn. Thanks for pulling up the live polls. Um, all over the place, um, but I see a big chunk of never. Let's see. Oh yeah. Oh, a big chunk of never and a few times a week, 26, 26%. OK, good to know. 1 every 4 person in this workshop, never, never used it, OK. So, we have a wide array, and if you look, what's also awesome about this audience is that it's from all over the world. So, um, curious to see how people use it all over. Um, so, um, let's go ahead. So we wanted to start with what is generative AI and um if we can move. Yeah, so it's a type of AI that creates new content. It's not the only type of AI. We also have other AI called like for example, nerve AI. It's a uh type of AI when you see in the uh Alexa, Siri. And all sorts of assistance, but generative AI creates content, creates music, image, video, computer code, websites, um, and so, so let me even back up, so yeah, so. The difference between AI and pre-AI is Historically, you had to. Uh, code or give commands to a computer. If the user does this, then do this output. Artificial intelligence learns, um, and it learns not by what was programmed, but as it learns through getting inputs over time. And so, AI driving a Tesla is a, is not generative AI as opposed to something that gives you an output like text or images or videos, and we're gonna be talking about generative AI today, yeah. So these are some examples. And if anyone's interested in anything that we brush over, like how to create art, writing stories, any of these things, we can talk to you during the event or after as well. Yeah, and and we brought a sample from our lab that we were experimenting. Without teaching you other languages, we were experimenting how we can make massive amount of educational material with you talking five different languages when you're doing your other job right now. This video is not created by Chat GPT. It is not chat GPT, but it is AI. So this is an AI generated video of, uh, it, it is actually me, but it's recreating me in a likeness of me, uh, teaching a topic that I. Um, asked Chat, asked Chat GPT to help teach, uh, a topic or maybe you took it from a script, uploaded it, and this is an example of gender of AI, not chat GPT, and how we're using this to educate the world. An inguinal hernia occurs when tissue such as part of the intestine protrudes through a weak spot in the abdominal muscles near the inguinal canal. Chiuhui fashion fukukoshanqi. To fundamental resos in chavista. So, um, it's not perfect because some people have noticed that the medical term incarceration when a hernia gets stuck, yeah, some of the pronunciations are not there yet because of the amount of training data and where is it coming from, so we don't have a lot of medical data out there to train AIs with the pronunciations because Uh, how many percentage of the all of the data all over the world is medical data, but it's evolving very fast, and we brought another example for, uh, AI video generation. The top of the Will Smith is eating a spaghetti last year. Um, you're gonna see it's, it's ridiculous. So the, the point of this video is how fast AI is moving, just how much things have progressed in the last year. We're gonna show you this. What we, what everyone says is AI is the worst it will ever be right now. It will be, and so we actually have to check each week because technology updates so fast. So this is an example of how things have progressed since last year. Look at the top video in the bottom. Oh, that's hot. That's hot. Uncle Phil, come. Yeah, try this. Both AI created, but look at the technology over the, how much it's gotten better over the past year. I mean, it looks completely real in just one year. So So what's this? This is, um, kind of a giving a, um, blueprint of how chat GPT works so you can get the best of it, uh, basically learning from examples and all the data, uh, whatever you put as input or questions, it breaks down the words instead of on the. Understanding it as a sentence, so it work, it looks at the word patterns and figuring out what you mean, and it gives you an answer from the experience and the knowledge it has. And, uh, with chat GPT's new, um, updates, it has memory now. Whatever you say, it updates itself, it updates the memory and it keeps getting better. Giving you personalized advice. OK. All right, so it's getting better over time. It'll start to get to know you in theory now. I asked, I don't know if this changed, and others in the audience may know, I actually asked Chat GPT. Are you getting to know me and learn me and it said no, we are not learning you yet and I don't know if that changed. I asked it that last year. We can see in the examples for uh some screenshots we brought, uh, whenever you say something new, it says memory updated, so it kind of gets an, um. Like list of things, maybe it's not getting like getting, getting to know you, but it's getting better what your interest in and putting those in a memory. And I think there's a privacy thing that they might say no, we're not trying to get to know you or we're not learning about you, but, uh, definitely keeping it all is in memory. OK, so a lot, this is the key slide I asked him to explain is. Um, the question that new, new users say to me is, why do I need chat GPT? I can look things up on Google. So we wanted to explain. Why a search engine is different than chat GPT and I know in general, we know they're different, but And by the way, Google is now has AI responses, but in general using a search engine, uh, gives you a list of, of websites to go to as opposed to aggregating. But, but, um, yeah, how would you explain the differences here? I think the most important one is the conversational interaction, which I think it's your favorite with chat GPT. You can, you can talk back and forth. You can share your feelings, opinions, ask for recommendations, and do all these things. It has the context awareness. So if you, uh, are talking. Talking about soccer and you keep you talking about soccer, it is gonna follow up the conversation and if you ask a question like 33 prompts down, uh, it's still gonna answer you because he knows what it knows what you're talking about and it gives you comprehensive answers and it summarizes for you instead of you have to sift through 15 different Google links and um, so, um, just regarding the conversational interaction, I want to impress upon those who wonder why I use it so much it's probably that, um. You almost, I almost never am satisfied with just the, the first time I ask it a question, it's the follow-up questions, and we're gonna keep showing you that. That's the power of using chat GPT is that you're asking it to explain something to you, and then you say, well, You know, imagine as if you were talking to someone in real life, rarely would you ask a single question and the conversation's over. It goes back and forth, and that's how I use chat GPT. Um, also, as I keep mentioning, chat GPT or Claude or Gemini, and we can explain what those are, they take everything that's out there and create a story for you, instead of just giving you a list of things. So it, it helps to aggregate an answer. We can use it for task automation, uh, to increase productivity. It can summarize your emails. We have, uh, some personalized GPTs that we're gonna share. It can write a letter of recommendation for you without you repeating the same steps every time. You just need to upload the CV and click enter. It's gonna give it to you. That's also a major way I use it. I've, I've offloaded a ton of work by having chat GPT do a lot of my routine work for me, and we'll show you some of those. Um, and tutoring learning, I know you use this a lot, but, um, we, we brought some examples from like all over the place. It's not just about medicine or innovation, um, like I, I, I'm into DIY, so I just ask some of the questions just like, can you teach me how to do this, because most of the tutorials, they assume you know some part of it, and if you're a complete beginner, it's a very helpful guide. And the other thing is If you're trying to learn something totally new, and you read an article or a book chapter, a lot of it doesn't explain at the level that you need. They'll use words you don't understand. With chachi PT, you can say, hold on, explain that at the level of a 9th grader. Hold on, go even more simple. Go to a 5th grader, and what is this word you keep mentioning it? I still don't get it, make this more simple. So it's great at tutoring and learning. Um, practice and feedback. OK. So, um, the other thing, just quickly, Um, there's a race for different chatbots. Chat GPT is from OpenAI, which is with Microsoft, or not really. I don't know. They're an investor. Yeah, it's, it's, it's. One of the founders is Elon Musk, right, but I mean that, and, um, they are currently, I think, working with Microsoft and also Apple announced that they're gonna integrate with Chat as well, so your series is gonna get an upgrade. I actually think Microsoft is winning this game right now because they have their own version, um, called Co-pilot, and they're also very much in partnership with OpenAI, so they're playing the field pretty well here. Um, but everyone's trying to get in the game. Apple has a huge announcement over the past couple of weeks that they're gonna be doing, uh, much better with AI, whereas Google. When you ask Google. Yeah. Well, my dog is sick and I wanna explain this to my 11 year old. Can you explain the acute kidney disease which Your dog had to 11 year old daughter, which you have and all Google can give you is what is kidney problems in dogs, dogs or any other like uh medical or informational things, but if you ask the same question to chat GPT, it tells you this is amazing. It tells you like how, like what is the level of an 11 year old would understand of a medical condition. And it kind of gives it in a heartfelt way. If this doesn't um explain, if this does not explain um Like the difference between a Google search and how to use chat GPT. This is like the best, the best explanation right here. Um, I'm seeing, um, yes, uh, Deb, we're gonna show the letter of support, that's towards the end. Um, and if people have to leave, we can get you the recording, I think. Uh, but we are gonna go through how we, we actually have a chat that we can send you that does it automatically for you, but we'll show you how we make letters, and then, um, Claudia said she used um recently to develop running training plan. And we're gonna even go through that as well. Um, and yeah, Microsoft was the first to make a deal with Chat GPT. All right, what about Siri? So they both respond to voice and text input. But chat GPT engages in conversations like we mentioned before. You can share your opinions, get recommendations, um, and the most important chat GBT will ask you questions in return. So it's a back and forth. And if you ask an open-ended question to chat GPT, you're gonna get a cohesive summary or a detailed text. But if you ask the same thing, uh, to Siri, uh, you're gonna get a list of web results, OK? And mostly Siri does the, um. Basic commands play music. Text my brother. Turn on lights, but I bet you by next year this slide will look totally different because Apple, it'll be different because they'll be working together. But like right now Chachi PT is generative AI and Siri is narrow AI. So narrow AI is trained in only very narrow. Area of uh the intelligence and they are just made to do certain tasks but it is getting an upgrade. OK, so, um, so you already said all that for example and you can feel free to do this if you want, um, and a broad example about how they sound. OK, so I love this part. So this, do you have to pay for this still is the in 40, I don't know, but I pay if you do the 0 is free. But does it have the audio or OK, I pay, I think it's 20 bucks a month because of this main feature. I love the conversational audio. So when I'm driving, when I'm working out, I literally Talk to chat GPT for my entire workout because I learned by conversation, and I'm never satisfied, so I might ask to summarize a book for me or read my emails, and I wanna ask questions back, like, well, wait, what was the main point there were? What were the tasks I need to know from that email? So I go back and forth. Um, with a book, I might say, um, tell me how this would apply in my work, I do this, and I can converse. So we just wanted to show you, I use chat GBT to read my emails and created me. What's called a GPT, so you can make your own if you use it a lot, um, and instead of retyping in the same prompts every time, you can actually create a templated prompt. So she made me one that summarizes my emails and you might say, why not just have Siri read your emails. I'm gonna show you the difference between how serious sounds when reading an email and then how chat G PT sounds. So this is, you didn't tell me you know Mike Tyson. Oh, right, right. So this, yeah, I know Mike Tyson. So this is just an email, you know, Mike Tyson emailed me asking me to teach him how to box and So, uh, I figured I would just, uh, play this email. You said you're busy. Yeah, I said I'm too busy. Yeah, I'm too busy, Mike. Sorry, yeah, so this is obviously a fake email, but this is if Mike Tyson were to send me an email, this is how Siri would read it to me while I'm driving. Hey Todd, hope this email finds you well. I wanted to reach out to you because I have a bit of a favor to ask. As you know, I've been in the boxing game for quite some time, but I think there's always. All right. Here's how chat GPT sounds when I ask it to read me my emails. Hey Todd, I hope this email finds you well. I wanted to reach out to you because I have a bit of a favor to ask. As you know, I've been in the boxing game for quite some time, but I think there's always room for improvement. And it, it's crazy cause it's getting better and better. So now they'll like stutter a little bit. They'll go, uh, uh, uh, I mean, uh, like they, it literally sounds like a real human and it's getting better every time. Um, I used to have the voice that sounded like Scarlett Johansson, but now they've removed it for legal reasons. Um, and also with the new 40, um, audio conversation update. It's gonna interrupt you when you're talking to them. Oh, right, so it's not, it's not just gonna wait for you to finish, it's gonna be like actual human, uh, like conversation that can interrupt you and tell you, uh, what you were thinking and vice versa. You can interrupt it and say, all right, I got it. Hold on. It's gonna be so much like a human conversation, and you probably will be talking chat GPT with more than some of your friends, probably, yeah, it's like the movie Her. Have you seen that movie? Yeah, OK. All right, um, all right, so. Everything's about the prompt. This is what M taught me. Um, a prompt is what you type in, and M's gonna talk about why everything is about, in fact, I went to classes on prompt engineering. That's like the new thing that people are doing. You can take online courses to get really good at getting the best answer by writing a really good prompt. So, so once you talk, you can advance if you want that. So, um, prompt is basically the, the question you provide to AI to get a response or generate information. Or, um, get the, to get the audio or image, whatever you want it to be created. The important thing is you need to provide context, you need to uh keep it as specific as possible to get the best result. Right, so if you're like, um, I wanna know how to, um, uh, best. Read a book to a child. The context would be, I am a preschool teacher. I have 7 kids in my class. I wanna, like that's the context. Giving it more about why you're asking the question really helps get you the answer you're looking for. If you give generic prompts or generic. Questions, it's not gonna help you, uh, a lot. That's some of the reasons people start using and stop using it, yeah, because they're just like, well, this is not that different than Google, right, because they didn't really prompt well. And in fact, you probably could create a GPT, that's like a template that explains who you are, so you don't have to say it every single time. Yeah, and we need a desired format for the output. We can say, uh, if you're looking for analytics, you can ask it to, uh, create as a table for you. You can ask it to be 100 words. If you can say like, let's do this like this and 200 characters because I'm gonna tweet this, uh, and it. Creates that out for you, yeah, OK, yep. And we have a small prompting guide, like we said, be clear and specific, and it's very important. Give Chat GPT a role. You can say, act like an editor, be my chef, these are my ingredients, uh, give me some recipes, or, um, you can break down the complex. Question or tasks into smaller simpler prompts instead of putting 10 sentences in one prompt, just ask uh step by step and question by question, and I think that's what is very important using audio version of chat GPT to go back and forth conversation. Yep, absolutely, yep. And like we said, provide context and examples, you know, if you're gonna say like, We would use it to, to take a uh a transcript and do something with it. We would first say, can you first summarize this transcript, Part two. OK, thank you for the summary. Now, can you tell me the tasks? So it's like, have it do it in a stepwise process and you'll get better results. So I'm gonna throw in an example that we didn't have in there cause I just thought of it. OK. And if anyone tells my wife that I use chat GPT for this, I'll be very unhappy. So. It was like our 20th anniversary back in November, and I was like, I asked Chad GBT I wanted to do something really special and romantic for my wife. What are some fun things I can do? And one of the things that mentioned is you could do a, uh, a scavenger hunt with her, a treasure hunt. I said, great. How does that work? And it said, well, you could um identify um Different stores that you would go to and, and buy her something and maybe have a clue to get you there. I said, great, can you help help pick? can you help me come up with the store, the store, the, the, the clues. I was like, um, here's, I'm buying her, you know, an earring at this store. Here's the name of the store, and it wrote the whole poem, a clue of a clue to help her get there. So it literally planned if I wanna be a part of this. I know. And so it planned the entire evening and then. Um, by the way, it planned our 20th anniversary trip where I said, I really wanna go to somewhere. Romantic. My wife likes wine. We wanna go somewhere in this month, so we want it to be not that cold, blah blah blah blah blah. And it planned our, and then through one question after the next, it planned probably one of the most special, um, trips we've ever had in our lives, and it picked, it picked a nice hotel for us and everything. So, and you got the credit, and I got the credit, and I was like, thanks, chat GPT, but you know, OK. OK. Um, so this is our first step of the day. It's, uh, explain chat GPT step by step what to do instead of one long prompt. And this is gonna be I think one of the helpful, most helpful things. Imagine you're talking to a real person, not a machine, not a computer. And when you keep it more genuine and casual, like you're talking to a person, you'll get the most, the better results. Be polite, clear, ask direct and specific questions for the best performance that you can get. There are some different, uh, techniques in prompt engineering, like, uh, when we set step by step prompts, that's called prompt chaining, and, uh, you can start with like list the steps in the scientific method, and then you can say, now explain each step of each step to me, and then you can ask, uh, this is when I wanna do the research. Can you explain how I can use this step by step approach in my research? Yeah, and I wanna, I'm just gonna stop for one second because I'm seeing a lot of discussion and comments about using this. In the hospital, if you're uh clinicians, um, be very careful. Uh, we do not do that. Um, because it's not, it's not HIPAA safe. It's, it's, it, if you upload it, theoretically, even uh someone could see it, so it's not HIPAA protected. If, in our hospital now, we use um a version of a chatbot called co-pilot, and it's totally safe. You can have it read your medical records, it can have, it can read the, the chart and summarize for you, it can do all that, but, and it's safe within the walls of the organization. Chat GPT is not, in fact, but if you wanna get help with how to treat a condition, um, Don't say that you have the illness because um it's gonna tell you to go see a doctor. It's gonna tell you to go see a doctor. So you say I'm a clinician. I have a patient with this. What do you suggest? And just take it with a grain of salt. It's not perfect. We're gonna show you how it makes a lot of mistakes sometimes, but use it as advice of, of a smart. Friend and always take it with a grain of salt. Or if you have a complex case, uses it as brainstorming, uh, friend that can put some, uh, possible diagnosis for you and, uh, you take the, the last step and just decide. Correct, you make the decision, yeah, sorry, keep going there. And like we said, that the role playing part, like, for example, you can say uh Act as a financial advisor and what advice would you give to someone who's new to investing? Probably they're gonna tell you, um, like when you ask how to invest on Google, you see pages and pages of information, but with chat GPT you can say, I have no idea about investing and I have like, let's say $500 that I wanna invest, how should I start? And it gives you a step by step approach. And if you say, You are a financial advisor or you're a retirement benefit specialist. It's, it'll give you different advice. Interesting. So you have to make sure you tell it who you want, who you are, but then say, here's who you are. You are a financial advisor, and it will give you a different result, OK. And um the, the constraining responses, you can just ask in 100 words or less, summarize the main points because sometimes that output is so long, it kind of uh Takes you out of your goal and takes you out of your path and we have of course limitations. One of them is lack of real understanding. It's basically does not understand what you say in a sentence or a paragraph. It recognizes the patterns and words. So if you put long complex prompts, it's not gonna understand because it doesn't have that real world knowledge. Second one is contextual limitations. Um, it has and, and knowledge boundaries. These two are like kind of similar because whatever they trained in, they know it. Whatever they did not have a training data, they have no idea. Sometimes they improvise, but how accurate those impro improvisations. So we need to make sure that either we need to fact check in chat GPT itself, and I would recommend, uh, whatever you take from chat GPT, uh, fact check somewhere else. Yeah, I mean, you'll, yeah, we'll show you. You can even ask chat GPT if it's lying. Um, that's one of your slides, right? You get into that, OK. And we have biases. Uh, it reflects training data biases, uh, like a lot of things in the world, uh, if there is a lot of research about certain condition or certain, uh, populations or whoever influences the science and poli politics, that's all data is loaded in the chat GPT, and, um, we also have confirmation bias coming from like all over the data, and this is very interesting. I'm gonna show you some pictures that Chat GPT created or maybe I'm sorry, another uh another AI agent mid journey created then it's asked uh by European perspective, how would people from different states of the US would look like? So this is crazy. So if you look, uh, there are like some certain like Georgia peach, and like Montana, like horses, Nevada casino, but If, if you look at them, they're all white, they're mostly men, and there are a lot of stereotypical things going on, yeah, yeah, so don't believe what you, what Chat GBT tells you. It's crazy. You really have to be careful of the biases, and I, I can just tell you that, um, that Sam Altman and who's the founder of OpenAI, they, they, they say that right now, there, there's so much opportunity for improvement. I think it's world changing, but even with that. This, this demonstrates the need for improvement. Yeah. Ask people to fact check themselves. Prompting the uh AI fact check themselves or if you ask, are you sure? Can you, can you check again? Can you do it again, it helps to uh redo the previous like scan the previous answer and it improves the accuracy and it actually trains itself in the meantime, so you would have more accurate responses in the future as well. Does this get into my, uh, yeah. So, if anyone tells my kids that this, I'll be really upset. So, yes, I cheated on Wordle. So, uh, I was going crazy cause I could find an answer to this. So I asked Chat GPT what five-letter word starts with A N and ends with T, and the only other vowel allowed is an O cause it was yellow, I couldn't figure it out. And it told me I'm not. And I'm like, that's not a word, and it was like, oh yeah, you're right, I apologize. So, it makes mistakes, it's going to give you an answer, and if you tell it, You can even say, don't give me an answer unless you're sure about it, and it, then it will give you a different answer. So, it will make mistakes. It's like a human. It is as human as you can imagine. So, just take it what, what it says with a grain of salt. I still got it wrong. It's what it still got it wrong the second time. Like it couldn't do it, but it's, it has a hard time giving up or just saying I don't know. And so it'll keep giving you answers I don't know and which is a problem they need to improve that. That's how you understand that's kind of acts like more like a human. And uh we ask which US presidents are from New York and they give you, give us like 78 different names, but some of them are not necessarily from New York, they're associated with New York. So let me say, can you please double check? I'm not sure if this is correct. The names, uh, go down and this time it provides you where was, where were they born. OK, so it just kind of tells you like sorry for my mistake, sorry for my mistake. I mean, it will always admit it's a mistake, but Yeah, one of the questions, uh, um, mistakes are not that common, uh, the obvious mistakes, so we had to go online and look for some of the mistakes. Someone asked which one is far from Earth, sun or moon, and we put the picture here if you have the same question. Uh, sun are definitely farther from, um, the earth compared to moon. It gives, it calculates it correctly, but it tells you. The moon is further if you look at the numbers and uh you just ask it again and it says clarify sorry for my mistake and this is the right answer but this is really rare. I know some people say, well, it could be wrong. Yeah, it could be wrong, but it could be any smart person you talk to, but for the most part it's very, very accurate. Just don't rely on it if it's something critical. If this is like a verbal example, like what are the names, the country names that says a Y, it says Zambia. Or like in less than 50 words explain this to me and it gives you a paragraph and you're just like how many words for that there's like 20 there's and if you ask, well that was not 20, we count again it's like whoops it was 24 but these are not that common. These are rare mistakes, yeah. So if you wanna take it away for conversational learning. OK, so, um, so what is this, um. So, conversational, this is basically um how you can use chachi PT to have like a conversation with it and teaching you, instead of just asking it one question. So this is, can you explain parallel parking in a clear and simple way? And here's what it says. It gives you sort of the step by step approach, the step by steps, yeah, position your car, reverse, whatever. Um, this is what, um, Google Google would show you, um. So here's, go ahead, this is yours. Yeah, I, uh, I was like I was, uh, learning Spanish online website and I have a problem understanding, uh, some of the grammar, grammar topics, and I asked Cha GPT and they're trying to explain it to me, uh, but with the results I still did not understand, and I said, well, I still don't understand. Can you teach me in a different way this time it changes the approach or maybe, um, gives you a like a simpler way or changes the approach. Uh, you can't even say, well, uh, how would you tell me if I was a 6th grader and stuff like that. So this, this is, I am said that's what I learned from you mostly. I love this. Like I almost always needed to explain it, you know, again another way. Just to make sure I totally am 100% understand it. To me, I love this function that, and, and, uh, you know, I don't have to worry about being embarrassed how many times, and I always apologize. Like I'm so sorry. Can you please explain it? But this time, and, and it's like no problem, I totally get it. Let's do it again. So it, it gives you a little motivation, yeah. That's very common not to understand. All right, tell this story. Yeah, like I, like I said, um, um, I, I like, uh, the IB projects, and we got a house last year, so we're trying to, um, at our touch in every different room, and I wanted to have this, uh, built-in, uh, fire insert wall. And uh all over the YouTube, there are a lot of tutorials and you kind of see people assume that you know a lot of it. They just use, show you how to assemble and stuff like that. And my main problem is, how do you build the frame that you're sure it's gonna hold the, the fire insert in the TV and all the weight. And I asked, uh, like, how can you, how can I start? Like, can you explain me with physics and architecture, and give me like nine-step approach, and I said that's a lot of steps. And I wanna start now. What is the first one I could do? So I gather your materials, do your measurements, and kept explaining, and sometimes, um, you just gotta try other techniques because it kept giving me like this is a step by step approach, still assume that I know how to frame. And I said, no. If you were the general contractor, and I'm very new to this, and I'm, I like, let's say I was your intern, how would you teach me distributing the load evenly so whatever I do, it's not gonna collapse and my wife is not gonna get me out of the house. So, um, I will try this and I'll probably keep people updated, but after this it was actually made sense, really. So this was the one that worked, yeah, it said like put like, uh, when you build your frame, uh, make sure that like you have every 16 inches. Because you have a piece of stud, use 2x4s, uh, make sure that like load bearing parts would be the, uh, like fireplace insert and the TV mount and make sure that you put like extra support to those places and it, it made, it made more sense. OK, I feel like I can do it. Do we have an example of the one where, um, I think you go into the one where how I say no, like how do I start today, is that, is that OK? So this is one of the things that you wanted, uh, is extracting necessary information from a long email or meeting. Before you keep going, first of all, I wanna thank a couple of other people. I see that Rami Shaban is answering a lot of these questions. Rami is also part of our team. Rami is a, uh, professor, uh, at Utah on instructional technology, and, uh, he is also a brilliant, uh, person on our team learning about AI as well. He's part of our AI team. And uh, so Rami is a great resource, so feel free, Rami, thanks for answering. And also, I, I think, um, uh, Cecilia Jeena is on the call. She's also part of our team, and she helped me give the same presentation uh in Vail, uh, and had some incredible ideas about how she went skiing in Vail and said, I'm a such and such level skier. I'm going to Vail. I'm living here. Tell me what. Tell me the plan out my day of what chairlifts and runs to do, and it did it perfectly for her. So, um, thanks everyone for helping here. All right, let's keep going. So we have this prompts. Um, we're gonna release this after the, uh, after the event as a PDF file. Oh, yeah, we can send them as a, uh, newsletter, but these are more generic ones. We brought some, um, it's the same email from Mike Tyson and So, I created this for me. So, Besides the fact that I like things read aloud to me like a child, um, I also like them summarized, and it was taking me a lot of time to do it manually, so I created a GPT that I, we can send to anybody who wants it. We'll share it in the next slide. We'll share it in the next slide with the QR code, but, um, look what it did. So I was like, all right, you know, this is an email where Mike Tyson's asking me to teach him how to box, and can you, you don't have to, you don't even have to say that. You can just, yeah, you can just paste your, uh, copy and paste your email text and say, um, especially with the, the our, um, GPT, you don't even have to say anything, copy and paste, and it will print out the same text. Don't be afraid, that's what it's supposed to do. Then you click, um, whole like. What, what would you say like lone touch read like, yeah, yeah, you click, yeah, you click on the text and say like uh pick read aloud, it will read the whole text to you and it'll ask you then, do you want to summarize action action items. So let me make sure I, and then there's a point I wanna make. OK, so yes, so it, the way I built it is it just literally. Reprints what I just pasted so that you can then hold and say read aloud cause I like that sound better and then it, she has it built so it immediately asks, now do you want a summary, now do you want action items and it keeps asking questions. But the point I wanna make is I learned this recently, I don't know if you knew this. If you push and hold on a response from chat GBT, it selects the entire thing. One of the choices is select text. So you don't have to select the entire thing. You can actually drag and, and, and, and just select the area you wanna copy. So yeah, I never use that one. Yeah, I always thought that like picking the whole thing and selecting is very too much cumbersome. Yeah, this is the read aloud chat GPT. Um, you can use the QR code and um we made it in a way that everyone has the link and use the QR code, can go ahead and use it. Uh, you, you can use this main buttons or just like copy and paste your text. Uh, it will work for you. So, so just to explain this again, what's so awesome about chat GBT is they're creating an open source sharing, so. Everyone is starting to make their own personal sort of templates that you would use over and over again. And there's a, there's a store that I can explain where you can go, it's all, almost all free, and just download them as a saved template on the, the left side of chat GPT. One of the ones that, um, I think, uh, Um, uh, one of my colleagues in Cincinnati sent me one that's like, is this gluten-free? That's it. So you just like type in such and such. No, it's not. Yes, it is. And I did that for another dietary thing called FODMA, like, is this FODMA? No. So you can create and share with each other, we could show you how to do that. All right. And this is the uh same email text. And this is the output. I wanna show this. If you're using on computer, you just need to hit speaker button to uh get the read aloud, and it will ask you, would you like text summarized with action items. You can say yes or no, but um. These are the actionable items from that GPT giving the QR code one more time. You couldn't get it yet. So, uh, the next one is how are you preparing for hard conversations or compose a challenging email which is gonna go with the demonstrating EQ which I know you, you like it. I love it. This is the part that is not intuitive. What I love about artificial intelligence is that it has great emotional intelligence, better than most humans. So, um, when I have a tricky email to send or a tricky conversation, It is so good at crafting it in an emotionally intelligent, tactful way. And I use that all the time. So if any of you have gotten really, uh, wonderful emails from me that are very emotionally intelligent, it was probably Chat GBT helping me do it. We send out, you can send out holiday emails, you can send out whatever it is, OK. And we brought some examples here, um, like, let, let's say your daughter came home and she's upset and she didn't get invited to a birthday party. How would you talk to her about this? and you can ask Chat GPT, how can I start the conversation with demonstrating EQ and, um, like be sensitive about like her, her, uh, sadness, and it will give you. Like very Therapist, like, it's really great, very, very conscious about what's going on, and it just tells you, well, I saw you were a little bit down today. Like you wanna share with me? It's so good instead of just like, what's going on? So it's so good. Look at, so there's one like lately my calendar has, has booked me out like a long time, and I always feel so horrible to like write, oh, I'm sorry, I'm booked out. 2 or 3 months, and then people think that I'm blowing them off. So I asked it, like, how do I respond in a way that doesn't sound obnoxious, but telling them that my calendar is busy. So it's like, thank you for reaching out. I really value your time and opportunity to meet. Unfortunately, my schedule is quite packed, so it's just much better than I am at saying it. Um. Or we can, we, we use it a lot for, can you rephrase this? Can you make this sound better? Can you make it sound better if you see here, I didn't like it. Sometimes I'll ask it for an answer like I need a word. I'm like give me a word that Would fit here would fit here and it gives me one. I'm like, no, give me 20 more because I don't like it. Um, what's this? This is about the brainstorming part, um. So, in our office, we're having an office space problem, and We have not enough desks for the number of people, and I said to the team, like, let's try to brainstorm a creative solution to the space problem we're having in the office, but we also can ask Chat GBT, here's our very challenging situation. Can you come up with a creative Out of the box solution for our problem. It doesn't even have to come up with a solution. Sometimes it asks you questions that you can come up with a solution, right, right. So it's, it's helping me brainstorm, yeah, um. This is another, um, idea. Like, let's say we have a project with a limited time. Uh, how can we find a solution to increase the productivity to meet the deadlines without, uh, doing like, uh, like keeping the workout same and the budget our same. So, uh, if it's something you can still say, well, can we brainstorm about this? Uh, I just don't want, like, if I don't like the options that gave me or if I wanna come up with more creative ones, I can even ask, can you ask me questions about the issue so you can understand it better and help me find solutions, right, right. I mean, honestly, you have to imagine this is like you're really smart. Assistant, who is like your partner in crime, who's just, you can be smart in almost any topic, but also emotionally intelligent, can brainstorm with you, can be creative, um, and you, you just, you, I would say like, if anything is like stressing you, if there's something you just are up at night thinking about, more than likely that's something you should try to see if this can help you resolve that. Um. This is about daily planning. Again, we have a like generalized um brief template, uh prompt template that we're gonna share after the event, uh, or you can um go as in detail as we can do. The first example is, uh, very basic. You can put it in Google as well. I need help with planning and prioritizing my morning routine. Can you provide some guidance, um, how to, I'm gonna organize my approach. It will give you, uh, like what are your goals, and if you look at here, it says memory updated so it knows what that is trying to implement the morning routine. Yeah. And it will pay attention to what you've said before. Or if you wanna go in more detail and more personalized, uh, this is what like you cannot do over Google or anywhere else. It is, uh, I wanna give you a list of everything I like to do in the mornings. I like to read a journal for 20 minutes, exercise for 30, uh, prepare, eat breakfast, I need to get my kids for school around 6:30. Yeah, I need to start working at 8 and do not want to wake up before 6. Build my morning routine just for me, and it gives you the morning routine, uh, like you can see that, um, your children need to leave around 6:30. It asks, it asks you to, uh, get the kids. It's ready at 7:30 and it was just like, well, can you double check the time we get for my kids getting ready is after actually they're supposed to leave, yeah, it redo this part and it's still wrong, but it's just working on it. It's just like a kind of back and forth conversation that it sometimes, um. I wanna see like how patient you are. Yeah, I, I would keep recognizing, we will look back and laugh next year, and look back how, how, how it is now and say, oh my God, it's like night and day better. So it's in my mind almost perfect, but there are occasional mistakes that will be fixed. Um, Yeah, if you want to implement something, you can ask, this is what I wanna do, uh, what should I need to do today to be on track with my goal in the next 10 days? And like again, like for, um, sticking to the new habits and implement changes, um, we have like a generic template here. So I don't know if we get to this, but if I'll just say it anyways, we can skip it. So I'm trying to build a new program, um, where we're evaluating outside technologies and, um. Part of that is I want to create a syndicate of hospitals that work together and I said, well, how can I go about Action I make, you know, activating this thing and start building this out. And it gave, like, you should do this, this, this, this, and this. It was a great sort of blueprint to how to do it. Then I said, OK, today is Wednesday. Tell me like what I actually can do today. Like what should be my first move today? Like, should I call someone? Should I, like, how do I start today? Then tell me what I should do sequentially over the next 10 days to get this activated, and it was perfect. It was like, first I would start off and call, blah blah blah. By tomorrow, you should plan on creating this. It's so good at continually refining what you're trying to do to get to an answer you need help with. Um, Let's go ahead. Yeah, you can, you can ask like step by step for any, any topic you want, right? Um, what is your process of right? Oh, so letter of recommendation. So this is, I, Claudia is on here, she, she knows that, um. Claudia, I, I asked her to help me write my letters of recommendation. I, if it's just the occasional, that's fine, but I actually get a, uh, I have to write a lot of them, and, um, it, it just was not, I could not get it done from a time standpoint. And so, um, and feel free to write in the comments if you think this is like inappropriate to do, but I think it's fine. So what I do is, um, If you really look at what a letter of recommendation is for most of the ones I write. They just wanna make sure that I give the nod that this person has met the criteria to be whatever. But then in the beginning and the end, I give very personal stories about how I personally worked with them. So if I needed a letter, I would have a lot. I always would go and write my own stuff, um, but here's what how I start. I need to write a letter. For, for MG. Uh, they are applying for a position at blah, blah, blah. Um, can you help me draft a compelling and professional letter that highlights their skills, achievements, and qualities. But then, You have to upload. So then I would say I'm gonna now upload their resume, cover letter, cover everything, and I also say first I'm gonna upload the request that I got to write the letter so you understand who's asking me and what they want. Please do not. You have to tell it, do not generate the response yet. I'm just uploading this, and then you'll upload. Good, good. I got it now. Go ahead and upload the resume. I'm like, all right, I'm gonna now upload the resume. Now Chachi BT. It's getting better, but it does have limitations on how much you can upload. Um, there was a lawyer friend of mine who wanted to upload a 300-page brief. Uh, it can't do that. Claude or Gemini might be able to. Claude is able to do it. Yeah, it's getting better, so you might have to do it in pieces, parts like just upload the first few pages, then upload the next few pages, and keep saying, please do not generate a response until I'm done, and then say. Here's how I know Em. I met her this way. She came here. I know her, blah blah blah blah blah. She is, uh, absolutely horrible. I don't recommend her, no, but like you can write anything. She's the best ever. She's good at this, and then it writes it all in a letter, and you can, um, and we, uh, and if you don't wanna deal with, uh, step by step, like writing all these prompts, we made the GPT for it. You just need to upload the documents, hit enter. It's gonna do everything for you, and it's gonna ask you, uh, we have an example here. Uh, this is my CV and says, uh, do you wanna focus on the personal qualities or achievements or do you want both? Uh, please provide any specific details or points. Uh, I said let's go ahead with the balanced one. It like wrote, like it started the letter like this for maybe like 5 paragraphs, it went like that. And again, after, after that, if you don't like it, um, you could just ask to, they can redo it or you can take this and personalize it. Yourself, it's always better to have a draft and work on it instead of writing something from scratch. It's a huge time saver. Uh, this is not necessarily, this is what we get from, uh, chat GPT and uploading everywhere and, uh, sending to the candidate, but this is something that you can work easily with your constrict time constraints. And then like Claudia can tell you in the chat, like I almost, we almost always revise it. She'll send it to me and I'll say no, for example, it might be someone that was just Someone you worked with a couple of times, you really don't know them that well. They ask you to write a letter, and it's like this person is the best ever. And, and so we write back, do not make it so superlative. Like they're good, but they're not all that, or be neutral, be neutral, yeah, or just say, no, this person is the best I've ever worked with. Please write it in that, uh, lens. And so you will have to play around with it, but I agree it's a good rough draft to start with. And if you wanna use the same GPT without, uh, trying to prompt by yourself, you can use the QR code here, and we'll probably include this in the newsletter, um, after the event. So another tip, which is like very uh popular lately, uh, you'll see if you get, uh, keep using cha GPT sometimes it is so long, the output. So long, so boring, and I can tell, uh, you probably that's the reason you the audio ones, probably. I don't wanna read all of that. So if you say to chat GPT after that long output, no yapping, limit prose, no fluff, don't be cringe, it also eliminates, uh, it makes it like out like out more precise, engaging a human-like, reduces the text amount. It also eliminates words like delve. Yeah, which you'll see a lot. Chachi Pti uses a lot. Yeah, that's delve into this, and that's what cringe, like, no yapping makes it shorter, and no, uh, cringe is like takes away words that people don't normally use in everyday conversation. Oh, by the way, chachi BT also likes to output in like bullets. It's like 123, and I'll also sometimes tell it, do not put this in a bulleted format. Write this out as a narrative instead of like a bulletin list, no problem, it does it, yeah. Oh, that was my anniversary trip, yeah. So this is a prompt that you can uh basically give. This is where I'm going. These are the, uh, bucket list items I wanna do, and this is where I stay. I think you did with the Charleston, um, there are like a couple places around the hotel, and you wanted to have a route, and it gives you, so just explain, so I put a video about this on social media, but Um, I've learned this from Cecilia, when she did that veil thing is I was like, all right. I'm starting at this hotel. I have 5 places I wanna go. I wanna walk it. Pat, tell me the route of how I should walk it, and it did the whole thing perfectly for me. So that was a great, uh, travel guide for me. And it tells you what are the, uh, the distances in between, where you should start, where you should end. 01 mistake I learned from this, I didn't tell it I want to end back at my hotel. Oh, so it ended me like 2 miles away. I was like, oops, I forgot to tell it. Please end me back at my hotel, so. And this is another one. Our family is going to Gatlinburg, and there's like 5 different options, and I'm like, we're only there for like 4 to 5 days. Uh, can you make an itinerary for, because, you know, like some of the theme park, like, and it, it actually recognized that you would spend the whole day in a theme park and you would spend a couple hours in this aquarium and then you can do go some do something else for the rest of the day. So it was very nice. That's awesome, yeah. And I, I, I don't know if you do this. Like, I don't know if you create a shopping list or a meal plan. I haven't tried it yet. I know, but yeah, um, have you, have you used it? I do, yeah, OK. So this is basically uh like if you want a meal plan or get a recipe out of Chat GPT you can take a picture of it and upload it to chat GPT or you can tell uh these are the ingredients I have, what is the recipe I can do. Let's say we have like ground chicken, broccoli, carrots, eggs, and maybe some cheese. It's really, it's really amazing. That's incredible. It really can take a picture of what you have now. I did use it this past week. I'm now have a new hobby. I don't know if I told you my, I like to make turkey jerky on my. own now. Oh really? Yeah, so I asked it, is it hard, so easy. You need a dehydrator, and I was like, So how do I make this? And then how do I make sure it's safe to eat? Like, does it kill all the back? And it's like, yep, dude. And actually that meal that was much better than the other recipes I had tried, so it was good. And this is, this is, uh, we're, we're going to your book club. Oh OK. Dead Poets Society I imagine you're talking to Chachi. Hi, I wanna, yeah, all right. Book Club. This is again, I keep saying everything's my favorite. This is one of my favorites. So, I really wanna read so many books because I've, I'm starting in a, a new profession, sort of, uh, in addition to surgery, but also in innovation, and there's so many leadership books or innovation books or venture capital books, and I can't read them all. So, what I do is I chat GPT it. So, I, I'm about to go work out, I put in my headphones, I turn on the audio version of chat GPT and I say, I want, and for the next bit of time, I'm gonna be discussing a book with you, and I say, this is the book. The obstacle is the Way, or whatever it is. Tell me the summary. But then for the next half an hour, as if it's a book club, I converse about the book and how it relates to me. Tell me more about this. Can you give me an example of how I would use that in my job if I had this uh email that came through, I was upset about it. Tell me how this book would address that. And I go on and on and on, and if I really love it, then I'll read the book. Oh, I don't read it, but I'll listen to the book at, at 2x speed. Um, but anyways, so that's, that's, uh, that's how I use the book club. And we have some examples here. You, you can, you can even ask like I've been hearing about this book. What are the like main concepts they include? And let's say you're interested in one of them. You can ask them to uh, give an example and how to like, uh, implement this in my life. And if you use the, uh, same chat GPT account all the time, it'll probably have more information about you. So it would give a more personalized answer or you would say, I'm I'm interested. And like stoicism or uh I'm doing an innovation job uh would this like apply to me and you can just keep doing like if I wanted to start today, what should be the first thing I should do and I know you love that question I love that question and I think it is very valuable because most of the time, uh, like everything else, it's a lot of things to process, especially if you're talking about like big books that are nonfiction, yeah, and um. So, I'll give you another example of how I use, so, if I want an efficient answer tailored to what I'm looking for, it's better than a Google search. I watched the movie again from a while ago called The Beautiful Mind with uh Russell Crowe, and basically, I think that the gist is that he had created game theory, I think that's what it was, yeah, he created game theory. I didn't understand what game theory was. I knew if I were to go on reading about it, it would take me a while to find something that would really teach it to me, so I said I just saw the movie. Beautiful Mind, can you explain to me game theory? And then it did, and I said, no, explain this part to me. It's like, give me an example of when I would use game theory in my job. Give me an example of when someone, when two people are doing something, give me an example of someone who did use game theory, then give me an example of someone who didn't, so I can compare and contrast. Um, also, when I watch a movie, if I don't understand the ending, I asked to, I had a long conversation about, you know, explain this to me. So, yeah, you can just delve into things, speaking of delve, OK. This is, this is another tip. Don't accept chat GPT's answer if you don't get it. Just keep asking. You have to have the back and forth conversations, and, uh, I think this was your quote. Don't forget this is not a query, it's a conversation. You're not trying to extract something from them. You're just like, uh, like whatever whatever you give as a feedback or another prompt, it also helps train themselves, so it's work for both ways, OK. Do you need help with writing cards or thoughtful letters? Yes, this is where a creative, uh, writing comes. Uh, you can ask, uh, I need your help as a creative writing assistant for occasion. You can put the occasion for a card or something. Um, This is one of the like my 2 year old niece, uh, just like listen to a dinosaur song and she's obsessed with it and I'm just like, can you come up with a bedtime story of a dinosaur that I can read under 3 minutes. I love that it's not gonna stay like it's not so it wrote you a story. It write me it wrote me a story and I said, can you rewrite this like according to Montessori principles? Wow, it is, yeah, it is like so popular like with the child education right now. Uh, I haven't read it to her yet, but. I love that. And I like, I sent it to you. I know you said it's very funny. Uh, we celebrate our cats got you Day, and they're just like, this is the 3rd birthday. I got him a card. I don't know what to put it in, and it gives him like awesome birthday wishes like celebrating you, meowing forever. This is my cat. I actually celebrate him, but he wanted to eat the cat, so. What's this one? So this is your example about like family vacation savings. Let's see. I don't know if my, uh, sister-in-law is on here. Um, Michelle Chu. Um, Michelle is our, uh, financial advisor. She'll be happy to hear that I did this. So I said, so my wife and I have three girls that are teenagers, 1113, and 17, and I'm trying to plan for the future when they all have families of their own. Uh, and I'd like to be able to take them on vacations. I'm trying to figure out the best financial way of doing that. Uh, should I put money away and take the interest or use it to spend on trips, blah, blah, blah. And it basically talked me through, and then you can see I did a follow-up in the top right corner, what would be the best destination for 3 to 4, uh, person, or for 3 to 4-person family trip. Or is that 3 to 4 trips, 3 to 4 family vacation? 03 to 4 families, I see. Yeah, yeah, yeah, 3 daughters, yeah, that part, OK, no, that would be a, yeah, um, so it's great, and I know M modified because I wrote it all wrong, oh, which was interesting. You probably used the audio. I used the audio and it came wrong, so I modified it, but it still understood what I was asking even though I said it all wrong, yeah. Knows you now. Yeah, I know. Um, oh, this was good. This was me not understanding something. I'm reading a sentence in a book called Corporate Venturing, and a man from Intel says in regards to equity investing, if our returns as CVCs exceed the weighted average cost of capital for our parent companies, we have effectively made money and also gained valuable intelligence about the larger ecosystem. And I wrote, what does this mean? And um, it explained it to me, and then I said, Um, this is another one, I think, yeah, uh, I'm, what's that? This is another example, yeah, I'm trying to understand what most academic medical center or university technology transfers office do and what do they not do. Uh, I just started wearing contacts. How can you tell if you're putting the contact in with the curve the correct way? Um, what was this one, what are the best books out there, and, and it gave me venture. I asked it for venture capital, and it gave me a bunch of recommendations, and I actually bought all of these books. Um, um, said, I'm just, I'm, I'm a surgeon. Like, can you give me a little bit more? I need basic, yeah, yeah, yeah. Um, could you start by asking me about times when I procrastinate? What's this? This is, this is one of the, um, you know, those buttons on chat GPT when you, um, turn it on. It sometimes you use to initiate the conversation. This is one of them. It just asks me to, I, I'm asking Chachi PT to ask me questions about what is my procrastination, uh, like time frame and how can I overcome it instead of just like giving me background information. I'm just saying, you know what, can you like, I don't even know what to tell you, but I wanna figure this out. Can you just ask me what is necessary to be known for you to figure it out. Um, here, there's a question about using this. Um, so, uh, so is Luzia, um, asked about, um, I would like to have it help me with content creation for professional LinkedIn. Um, there are so what I would recommend is if you go on TikTok and type in AI for social media content, that is probably one of the most prolific videos out there. There are so many great tools, but what you would do, uh, Lucia, is you would say, I want to. Um, create, oh, this is, oh, I'm so glad you brought this up, cause we did this with my wife, Diana. You could say, I wanna create my profile for LinkedIn. I am such and such. Can you help me? But let's say you wanna put content out every day of the week. You wanna write blogs. My wife is a facial plastic surgeon, so she said, I wanna write blogs. Can you give me a list of things that we could write about? And I gave her 100 things. It's like, One day write about this, then write about this, and you could even have it help you write the blogs. Um, it could literally be your entire social media content creator. You just have to make sure it's accurate and it's correct and give it more about you, um, give it documents, upload anything so it really knows what you are, and then it will help you create content. Also, they can, uh, they can ask Chachipiti to ask, uh, act as a. Uh, experienced social media manager and come up with the topics, it will be better than just asking what are the topics I can write about. Yeah, exactly. If you tell it who it is, there are so, but you gotta look at these videos. There's so much, uh, that's what a lot of people are using this for, Lucia. So I would look online on TikTok, and I, that's what I follow. I watch them every day, videos about how people are using this in their job, and social media is great, yep, um, TikTok videos too, like literally say. I wanna make videos. I wanna do this. Give me some ideas for content and then tell me what to say in the video, and it will tell you everything, yeah, yeah. And uh that you can, it's, you can get quizzed by Chat GPT if you wanna learn something. This is one of the examples, um, when you, when you do, uh, when you get it wrong, it says actually that's not the right answer, but let's keep going. I like the effort. It doesn't, uh, beat you down when you get it wrong. I like that, so you can make a whole Trivia Pursuit game on, yeah, OK, you can, you can ask like, uh, you can like you, you can even say like, oh, I'm interested like. Uh, like a baseball, uh, and I wanna learn more about it, and, uh, can you like ask me questions about the players, uh, about like trivia stuff, uh, it just, I don't know, it's just like you can, uh, you can play. So do you get into for this, so just in case, are you getting into, uh, image generation? No, OK, so in Chat GPT, the paid version, there's Dolly and I said, can you create me an image of such and such, and it makes it's, it's amazing. We can try to maybe um send you those later, but you should use the image generation and chat GPT for the paid version because it's, it's amazing. So if you're giving a talk and you need slides, it can make the image that you want. Yeah, you can even ask. I just moved a new city. Can you help me make friends? Oh really? Yeah, what should I do? It sounds like you can be volunteer. You can go to local events, um, like you, you can just like go to a library event or like it's stuff like that, or, um, like. I wanna like um touch base with that like content creation part uh if you even ask, uh, it might, it, it can put it on a table for you for every day on a content calendar as well if you wanna tell them like, um, can you give me the results for um like. Rows are days, columns are, um, week of the month. Can you give me a calendar for June 2024? So and we'll just fill it out for you. It honestly, I would push it to the limits and it's getting better. You can have it analyze data. You can upload an Excel spreadsheet and say, tell me, you know, what this means, what, what, how can I interpret these data? Um, can you, I, there's even some that can create graphs now, by the way. You can make your own websites now. Um, oh, you're gonna, oh my God, that was recent. OK, go ahead. You, you can ask Chat if you could create a personal web page for me all in a single file. Ask me three questions about what you need to know, and, um, it asked questions about like what kind of sections you wanted, and I said I want about me. I want a portfolio and a contact page. Wrote the quote for me. And um I said like can I see how it looks Chachi PT is not able to do that yet, but it tells you how you can see it. It says like copy this code and um they can put it in your uh browser or like save as a document and you can uh look into it. But I, I heard from you Claude can show you a little demo as of last week. Claude will show you the demo. This is game changing, guys, so. There's a lot of people using for coding. Software developers are, are in trouble a little bit because, um, you literally can describe exactly what you want your app or your website to look like, and in a second it will print out the code for you and you've just made your app. So, or they can upload their codes and ask, this is not working. Where's the problem. Oh, it can diagnose. Yeah, it can diagnose it as well, and I think that's one of the bigger pains in the, uh, the software development part because they have thousands of lines of code. You don't know which one is it would take forever, and that just finds it. It's like this is the mistake. This is, I mean, I think software development is gonna be game changing. This will be one of the biggest, um, use cases of chat GPT, um, just to give you another example. Drug development. The way it works is they throw millions of particles at a protein to see what sticks. But now you can actually ask Chat GPT what particle would bind to this protein, and it will help produce drugs. It's, it's incredible. Um, all right, let's see what else. That was it. That's it. We got through it in only, yeah, we have 45 minutes, yeah, so, um, let's put this poll up there after this workshop. How comfortable do you feel using chat GPT? Um, not comfortable, somewhat comfortable, uh, makes more sense now or very comfortable. Um, let's see. Oh, it makes more sense now. Yeah, exactly. So we, we, we don't know how this will compare to, uh, look, there's no not comfortable. No A's. All right, that's impressive. It is impressive. Wow, you did a lot of people are very comfortable. Um, Considering 1 of every 4 people did not use at all, that's incredible. I, so, um, we can end anytime people want us to, but honestly, we're here now to answer any specific questions that people may have. There was one, that, um, came from, uh, Daniel. He said, thank you, this is wonderful. Could you please let me know if you designed the QR codes? There's a million QR code creators out there, um. But I can tell you how she we use the QR code generator. Yeah, so I have two on my iPhone. They're free. I generate QR codes all the time. You can get those on the app store. We designed the GPTs, the GPTs M creates, um, it, it, um. If you wanna know like what's the gist on how to make a GPT. So, uh, if they have the paid version of chat GPT, there's a section, uh, it says GPTs, and you can go and say add new GPT GPT or customize it, and it asks you, uh, basically like, can you walk me through step by step what you need this for? What should I know about this and what do you want this GPT to do? OK. And then you can even, um, upload a lot of documents as a, uh, library for that GPT so you wouldn't have to upload the like this is so good for task automation. If you, if they know what they need to do, uh, and if they have the library and the memory for it. You can just copy and paste like we did with the emails and the letter of recommendation instead of asking the same thing over and over again, but it's, it's pretty easy and there's a lot of tutorials out there we can, uh, maybe share some screenshots on, um, like a PDF document after the event event for, uh, people who wanna, uh, try it, but I think you need to have a paid account for that. Um, are my requests saved? Is that Romney says in the paid version. I, I don't actually know what's, um. In the free version anymore, cause I've had the paid one for so long, um. So, again, let me know, even if you ask your question above, let me know now, um, if there's any questions that we can answer. If not, we will, um, oh, Claudia says she uploaded Smartsheets to create a more detailed GN chart. I mean, honestly, it can, it can do almost anything, and if not, there are plug-ins you could probably buy the, the GPT store. It is incredible. I mean, I saw ones where you can like try to find a, a, a name of a song if you know the, like there's so much you could do that's out there, um, but I think we can probably, um, end early if there's no other questions. Um, but, uh, um, thanks for and did all this work, she put this entire thing together. Um, if you couldn't tell, uh, she, she, she clearly knows her stuff, and I'm, uh, riding her coattails. But thanks, everybody. Let us know if there's something you'd want in a subsequent one, if there's anything we didn't answer here, but, uh, uh, I guess with that, uh, thanks, Carolyn and Ayla. And, uh, with that, we'll say good morning, good afternoon, good evening, and we'll see you next time.
Click "Show Transcript" to view the full transcription (72233 characters)
Comments