How to Use AI Models for Data Analysis (Qual & Quant) (Full Transcript)

[00:00:00] Speaker 1: Hi. Hello, everyone. If you're interested to learn how you can use AI models such as ChatGPT to do data analysis, then you're in the right place. My name is Anmoye Brown, and I have been using AI models for evaluation purposes, and I'm joined by my colleague, Dr. Philip Adu, who has published several books on qualitative analysis and he is also an expert on the use of AI models. Dr. Adu, welcome. You're the right person to ask, and I must start out by saying, I thought ChatGPT and other AI models were for basic queries, but are you telling me you can do data analysis?

[00:00:53] Speaker 2: Yes, you can analyze your qualitative data and also quantitative data. At first, when it came, I was using it for qualitative analysis, trying to identify themes and also extract significant information from the database on a research question, create tables. But now you can also analyze quantitative data. You can run regression analysis, descriptive statistics, and you can also do visual representation. It's an open opportunity for us. For me, for example, I'm more of qualitative than quantitative, but using this tool, I'm able to do complex analysis and that are quantitative.

[00:01:41] Speaker 1: Thank you. You heard it well, the person's just joining us. I see Ate, I hope I'm pronouncing that correctly. Welcome. Please drop in the chat where you are listening from. It would be nice to know the geographical representation. For persons just joining us, you heard correctly. Yes, you can use ChatGPT now for quantitative data analysis, running all those statistical tests and so forth. Since I've spoken about that, Dr. Adum, for someone who is new, because all of this is new, quantitative data analysis and so forth, what tips do you have for someone who is new to using AI tools for data analysis?

[00:02:29] Speaker 2: I think one thing that you have to do is to try to explore. It's all about exploring and knowing the kinds of questions that you want to ask the system and to get information from there. I think that first thing that you have to think about is, the more the system know about your data, the better the results. You have to provide a little background information. Before you ask the system to analyze your qualitative or quantitative data, why don't you give a little bit of background information, the purpose of your study, how data was collected. Sometimes you ask the system, do you have any question for me to clarify before I ask you what to do for me? Allowing the system to ask you questions about your data before the system analyze for you is the best approach. Also, it can give you wrong information, so you have to be very careful. You have to also ask the system to do some self-reflection. You can say that, can you review what you are giving me to make sure that I'm really addressing a research question, or is there anything that's missing that you didn't look at? Asking the system to reflect upon the output is also another best practice. Just have interaction, be clear, provide the system all the information that it needs for it to help you to make sense of your data.

[00:04:00] Speaker 1: Very clear. These are just some of the tips. I got in there that you have to have clear prompts. Prompts is just queries. That's a nice way of saying, asking AI, we say prompts. You have to be very clear. These are some of your tips if you're new and you're just beginning on data using AI models for data analysis. I'm looking at the chat and I see we have persons from Ghana, Somalia, Canada, Baltimore, so quite a geographic spread. Please, we want this to be interactive. If you have a question, drop it in your chat so that we can see it and respond to it. Dr. Adu, you spoke of best practices. What about tools? What are some of the AI tools you can use for data analysis?

[00:04:55] Speaker 2: There are a lot of tools around. I'm going to also share with you what I've been using and what is helping me. Let me quickly share my screen and then we go through some of the tools. If you have any question, I'll be happy to provide you more information. Let's go to this PowerPoint slides. I think that will be helpful for us to talk about the tools. The first tool that you all know is ChatGPT. This is the tool that you can start if you are new, just to ask the system some questions and then get some responses. You can treat the system as like you are interviewing somebody. You have to be clear, simple questions to get information from the system. If let's say you have a qualitative data or quantitative data, you can use ChatGPT, you can upload your transcript or your data and ask the system questions. But as I said before, you ask the system question, make sure that you are provided a little bit of context about your study so that you get rich information from the system. Another one that is so useful and even getting better than ChatGPT is Cloud AI. I know that some countries you don't have access to this software or the AI tools, but those who have access, you can try it out and see what you're going to get. It's so powerful. It can give you visual representation about your data. It can also give you interactive visuals where you can click on and move the visualization. If we have time, I can also show you some of the things that I've done. If you have qualitative, quantitative data, you'll be able to upload and ask the system some questions and you'll be able to get information from it. Also, I really like Julia's AI because what it does is, it's like a platform that has a lot of AI tools there. It has ChatGPT, it has Cloud AI, it has other tools there that you can decide, okay, for this data, I want to use ChatGPT. For this data, I want to use Cloud AI. For this data, I want to use maybe Meta or any other software or tools that are in there. It's like a platform that has a lot of tools. You just choose the one that you want to use and then run your analysis. It's mostly good when you have a huge data, quantitative data that you want to make sense of. This one will be very helpful for you. The last one that is so interesting because we all know the qualitative software that are around in Vivo, Mask, QDA, Atlas AI. The good thing about Atlas AI is that they have introduced AI function in there, where let's say you are not sure about the things that you want to come up with. They have a function called Intentional AI Coding. When you go there, you tell the system, this is the purpose of my study. The system will suggest the research questions for you. You can adjust the research question based on the suggestions. Then it will go through your data and develop codes and themes for you. You see how now qualitative data, qualitative analysis tools they are using. Also, they are incorporating AI into the analysis so that you'll be able to make sense of your data within the shortest possible time and also get very insightful information from the data that you have. These are the four tools that I have for you. If you have any questions, as I said, there are a lot of tools. But these are the ones that I've used and then I'm sharing them to you.

[00:08:58] Speaker 1: Great. There are some questions coming in, Dr. Adu. But I have an activity for persons listening. Dr. Adu says he might show us some visualizations and what he has done with data. If you want him to do this, please just type yes in the comments or in the chat. We are here to serve. If you want him to do this, just click yes. I have some questions coming in and this one is from Sharon. Sharon asks, how do you mitigate against hallucinations, especially with qualitative data analysis?

[00:09:46] Speaker 2: There are some processes or best practices that based on my experience and based on what I've heard from my colleagues. I think that first of all, you have to be skeptical about any information that you are getting from chat GPT or any AI tools. Because they have even stated that the system can provide you false information. I think that the same way, sometimes we are skeptical receiving information from others. Asking the system right questions will help. First of all, you have to provide context. Second of all, you have to provide clarity. Make sure that you are asking a clear question. You have your research question. You can say that, this is my purpose of my study. This is my research question. Can you go through the data, extract information that are significant? You can also provide examples. For example, this information is significant based on my research question. Providing examples also helping the system to provide rich information. Then when you get the results, you can say that, okay, now you have given me the things. Can you go through the data and give me evidence from the data I've given you to show that these things are really from the data? Continue questioning the system to reflect on the findings will also help to get rich information. You see how you can ask this. Another thing that you can also think about, you can do a little bit of triangulation. Using more than one AI tool. You can use ChargePT, ask the same question, get the findings. You can use Cloud AI or any AI tool, compare and contrast. That's another way of making sure that you are getting rich information. But I think that more of questioning the findings, try to ask the system to provide evidence from the data to support the information is the best way of reducing hallucination.

[00:11:46] Speaker 1: Okay, great. Great one. And I see a question came in asking for clarification. What is hallucinations?

[00:11:54] Speaker 2: Okay. Yes. Oh, can you define it since? Yes, you are already asking, right?

[00:12:02] Speaker 1: Okay. What is hallucinations? I should have said that before. Hallucinations, you know, even outside of AI context, is when you see things that are not there and you're making up things that are not there. So when it comes to AI, we don't say AI is lying or making up things. It's just a fancy way to say the AI is hallucinating. It's having hallucinations. It's giving information that does not exist. And I think the takeaway from what Dr. Adu says is just like with human research, you have to triangulate, you have to validate, that you don't take any analysis. You have to verify, confirm, cross-check, validate. And it's most important as well when you're using AI. Dr. Adu, how did I do?

[00:12:55] Speaker 2: You did perfectly well, right? So you talk about, you know, the system confidently giving you wrong information. And one example is that in the system, you can tell the system, oh, can you go through this transcript and extract information that are significant for me? And it will extract that information. Some of the information may not be from the data that you gave to the system, right? So in order to be sure, what do you do? You search the information from the transcript and see, oh, is it really from the transcript, right? If it's not from the transcript, you tell the system that I've realized that most of the information that you gave me is not from the transcript. Can you only bring me information solely from the transcript I've presented to you, right? So letting the person or the system know that, okay, you have made a mistake. Please go back and look at it and then bring me the information that I need. You know, it's all like, this is where we always say that AI, they are assistant, they are assisting you, they are not leading the research. They are helping you to reach your destination. So you have to have some kind of control and supervision, questioning the result so that you get rich information from the system. So that's what you have to do. So how elucidation is, the system confidently giving you wrong information, right? So you always have to be careful about what you receive and, you know, question the information.

[00:14:24] Speaker 1: Indeed. And for persons just joining us, we're speaking today on how to use AI models for data analysis. Dr. Odu just gave us tools. Could you quickly recap those tools? I saw ChatGPT on there. What were the others?

[00:14:41] Speaker 2: So we have the Atlas TI, which is a qualitative software, but we have incorporated AI tools in that. We have Cloud AI. We have Julius AI. These are the tools that you can use to analyze your quantitative and qualitative data for. Atlas is for qualitative, but the rest you can analyze for both.

[00:15:02] Speaker 1: All right. So the question we have now, you have given us options, but a question from someone listening is, how do you choose which tool to use based on your purpose?

[00:15:13] Speaker 2: It is same thing as you want somebody to come and do a repair in your house, right? In order to get the right person, you call maybe two or three people and come and look at the thing and give you estimate, right? So not all AI tools are equal. It all depends on what you analyze. That's why I like that you have to explore. I cannot say that, oh, ChatGPT is perfect. It depends on the data I'm giving to the system and what I got from the system. Maybe Cloud AI or Julius will be perfect for you. So I always say that choose one or two or three AI tools, familiarize yourself with the system, right? And then use the same questions and see the kind of result that you're going to get. And there's a website that you can compare AI tools. That's what I like. So I'm going to put this, let me see what I can show that. And then you can... While Dr.

[00:16:14] Speaker 1: Odu looks for that, while you look for that, I have a contribution to make. And I use two AI tools. I use ChatGPT and I use Perplexity.ai. Oh, three actually. ChatGPT, Perplexity.ai and Cloud.ai, but I only pay for ChatGPT and I use all three. And I deliberately compared results. And over time, I get a sense of which tool gives me better output, depending on what I need. So like ChatGPT, I use that model less for my social media posts, because it's too sensational. I don't like the writing style. Cloud.ai, I use more for report writing and data analysis. So as Dr. Odu says, just experiment, try to have one or two, use both, and over time, you can get a sense of which is better for what type of analysis. Dr. Odu, do you have the comparison tool ready to share with us?

[00:17:15] Speaker 2: Yes, I have it. I was trying to share my screen, but it didn't work out. So what I'm going to do is to tell you, it's called LMSSYS Platform. So let me put it in the chat box. I was having difficulty sharing. So let me see whether I can first put it for you, Anne. And then you can share with them. And also, you can always email me if you want to, so that I can provide that information. But let me see what I can put in the chat box. Anyway, so it has a lot of AI tools. You just choose the ones that you are interested in, explore, and ask the system a question. And then the ones that are very good for you, you just choose. It's called LMSSYS.org, right? So when you go there, you'll be able to compare AI tools. If you are not getting it, you can email me or you'll contact me through the social media platform, and I'll be able to provide you that.

[00:18:21] Speaker 1: We will definitely put it in the chat for some reason that functionality is not working from our side. So Dr. Odu, I got a lot of yeses for the visualization. But is it that you're able to share your screen now?

[00:18:34] Speaker 2: Yeah, I was trying to share, but it's showing that unable to share screen. I don't know why. Oh, let me see. Okay, so now, okay. I think I'd be able to share. Let me try again.

[00:18:50] Speaker 1: Let's hope so, because we got a lot of yeses.

[00:18:54] Speaker 2: Okay, so let me go to... Let me see what I was able to share.

[00:19:03] Speaker 1: Yes, I can see you. We see you live.

[00:19:07] Speaker 2: Yes, so perfect. So can you see Cloud AI? I don't know. Maybe some of you don't have access to that, but it's so powerful, right? And let me share with you what I did now. Because of time, I cannot just give you the prompt, but I just want to share with you the information. And then maybe one day I can put it on my YouTube channel, the processes. So let me go up a little bit. Let me see what I'll be able to go. So I have a data about mental health, right? Mental health stigma. 100 participants were interviewed, right? And they were not interviewed. They completed a survey, 100 participants, to find out their perception about mental health stigma, and also their views or their cultural intelligence levels, right? So I uploaded the information to this system and I asked the system to review the data and give me general charts and graphs. And then so when I did that, this is what I got. So you see that I was able to... Dr.

[00:20:18] Speaker 1: Adu, I think your screen is showing Julius. Oh, really? Yes. Oh, I'm so sorry. Thanks to the user who alerted us.

[00:20:28] Speaker 2: Thank you for that. Okay. So, you know, this is about technology sometimes. So let me see whether I can... Yes, I think now you can see Cloud AI, right?

[00:20:42] Speaker 1: Yes, and I see some bar charts.

[00:20:44] Speaker 2: Yes. So this is the data that, you know, I uploaded, right? And then I asked the system generate chart and graphs for the demographics. And then the system has generated graphs for me here, right? So when you scroll down, you can see... So let's say you want the system to give you information about demographic results. You can ask the system that you'll be able to do that and then provide a chart on your right side. That's a good thing. And it also provides you quotes, which I don't have any experience. So I don't look at that place. I just look at preview to see. And then I went further and said, can you show a graph about the interaction between education experience and also cultural intelligence? And then the system was able to provide a graph for me. So this is the distribution about education, the connection. So you see how... Let's say you are doing a presentation and you need this information so fast. You can ask the system and you get this graph and you'll be able to show to people the connection between education experience and also cultural intelligence based on the data that I have, right? And also, if you want to have access to the data, you can contact me. I'll send you the data so that you can do the same thing that I did, right? So you see how powerful just asking the system questions and it will be able to provide you graphs. It can also explain it to you for you to see. The good thing is that you don't have to have a strong expertise in quantitative analysis, right? Just a basic one. The system can give you the results and then you can ask if you don't... Let's say you don't understand this information. You can say, can you explain to me what this graph is all about? And the system can break things down to you. It's just like your teacher teaching you or giving you all the information that you need, right? So you see that this kind of system empowers you instead of looking for somebody to help you to analyze your quantitative data, explain to you step-by-step how he or she did it, right? This system can provide it to you and then you can also ask the system question if you don't understand. It's just like your friend, your companion. So that's so powerful. And as I said, I don't have a huge background in quantitative, but this one has really empowered me to do complex analysis. And then I was able to do a regression analysis. And then you can also develop an interactive, what do we call it? Graph where you can manipulate. Let me see whether I have this one. So you see here, I can manipulate. I develop a graph. Imagine that you are showing people about the data that they gave you and how things might change if you move the regression line, right? You see how you can easily, this one is just a prompt. I just said, make the regression lie adjustable. And then the system was able to provide this to me. You can even add sound to it, sound. Let's say whenever you move it, it makes sound. So I don't know whether you can hear this sound. I don't know whether you can hear it. So it's so powerful to, I wish everybody has this access to this one. And also I was doing a little experiment here. I gave this app AI to my daughter who is 13 years old, right? To use it. And she was able to develop, what do we call it? A game, right? I think it's, I don't know the name. The one that's when you develop a game where the blocks are coming from on top and then comes out. Yes, she just asked the system and the system provide develop again. And she was able to play the game here. So now you don't have to learn about code or to learn to develop a game. You just ask the system a question and then interact. And then you create something, right? So you see how, as I said, it's empowering us. It's not gonna take over your job but it's gonna empower you to do complex things that you may not, you should have gone to do a master's degree to be able to do it or PhD. But now you can, with these tools, you can do many, many things, right? It's all about exploring and learning more and using it in a way that is going to be useful for you. So this is what I have for you. Because of time, I cannot go into more detail.

[00:25:35] Speaker 1: I wish, let me see, let me see what I- Let's do, you know, I want to get because you showed quantitative aspects but Benny had a question here on qualitative analysis, data analysis. And it is, how can you use AI to find themes and create a code book from- It's so simple.

[00:25:58] Speaker 2: I have videos, when you go to my YouTube channel if you type Philip, I do, I have a lot of video of using AI to do the qualitative analysis, right? So you can just ask the system. So you can say that, can you review the transcript and go through and identify information that are significant and develop themes for me addressing this research question? And the system will be able to do that. If you are not satisfied, you can say that, can you make an adjustment to the themes? And you can give the system an example so that the system will learn from your example. So it's so, so powerful that, yes you can use it for qualitative analysis but people are a little bit concerned because they think that, okay, is, is the, what if will people believe the findings from AI? And I always say that it's all about being transparent telling the people about the steps that you followed to be able to reach your, the theme that you have, right? So you have to really think through when you are using the system, try to document the process so that in case somebody asks you how were you able to develop code using AI to be able to tell the system, okay these are the terms that I use and this is what I got, right? So being transparent and also showing them how you verify the results is also a very way that you make people believe what you found. So this is a tool that my daughter did it. So you can see that she was able to use it and play. She just asked questions, few questions. I can show you some of the questions that she asked. She said, can you make a game like Tetris game? And then the system did this one and then you see, it's not going to be perfect. So she asked more questions. Can you make it so that I can play now? And the system was able to make it so that she can play it on the screen now. And then can you, you know, just continue to ask questions and then you'll be able to get information, right? So this is not what I did, my daughter did it.

[00:28:04] Speaker 1: So great, so we see not just for data analysis but for other things, coding that you don't necessarily have to have a background in coding but you touched on something very important. But for everyone listening in, just say your yes in the chat if you are still with us, if you're still following if this is still interesting, just put yes in the chat for persons just joining in, we're discussing how we're using AI tools for data analysis. We learned that you can now use chat GPT for quantitative data analysis to do sophisticated quantitative tests, multiple regression, chi-square and the sort of things. Now, this brings us to a very important question that we have from someone here asking, all this data that we're uploading, how is data protected and enforced in these AI models? And we're talking about ethics as well. Is it wise to upload data?

[00:29:04] Speaker 2: Yeah, so you always know that companies will promise you that they're going to protect your data and you shouldn't always rely on a promise because they want you to use their tool. They are investing so much money, billions of dollars. So they have to entice you to use their tools. But we also have to be very responsible, right? The same way, the traditional way, you have to take off all identifiable information before you upload anything into their system. You also have to have time and ask questions about how they use your data, right? So GPT OpenAI may use your data to improve their model, but you have an option to opt out, right? There are many ways of opting out. When you go to your settings, you can say that, okay, don't use my data to train your system, right? Or you can have a temporary chat. I can show you where to get that. Let me quickly show you. Yep. So let me pull this one here. So when you go to your account, you see temporary chat. If you don't want the system to remember or save any conversation, you can use temporary chat to have a conversation with the system, especially something that is sensitive that we don't want the system to have. But if you want the system to, you don't want a system to use your data to train, you can go to your account. Let me see here. Let me go to your account. And then I think you go to settings and then you can say there's a place about data control and then you can turn this one off, right? That you don't want your data to be used to improve your system. So there are mechanisms to control your data, but I think it's not foolproof. What you have to do is you just have to be responsible. If you have a company sensitive information, you should not use it in the system because you don't know when it will come out or maybe their system might be hacked and somebody get access to that information, gonna be in trouble, right? So we just have to be very careful. Don't use any AI tool that you are not sure because remember the everyday new AI tools come, right? And they want your email address. They want you to upload information. Sometimes they use it to train your system. So we have to be very careful. Another thing is that the AI companies are reaching a stage where they are not having enough data generated by human beings to train your system. So now they are relying on, we call it synthetic data, allowing the AI to generate data so that they can use it to train your system. So now data is becoming like gold, right? So every AI company wants your data to train your system, right? So you always have to be very careful which kind of AI tools that you use. So that's why I use the common one, open AI one and other ones that I feel comfortable. So if you are not sure about it, please don't upload your information to the system, right?

[00:32:44] Speaker 1: Okay, great. And please use the chat to type any questions that you might have and take this time to click like and share. If you like what you're hearing, click the like as well so other persons can know and click that follow button so you can follow Dr. Adu and I because we'll be having more of these live streams. So you'll be notified when we have our free events. So next question, Dr. Adu, for someone who wants to learn more and interested, are there any courses, any programs you could recommend that they could take to learn more about data analysis?

[00:33:24] Speaker 2: For that, you can go to my YouTube channel. I have a lot of videos on, I can quickly share my YouTube channel. I have a lot of videos on that and I try to make sure that I teach my followers how to use the system in an ethical way, right? I don't just jump or new AI tool can do this but how can you use in a way that the information will be credible and that's what I always focus on and do. So when you go to my AI, my channel, where is it? Yes, so if you search Philip Adu, this is my channel, you'll be able to get access to, I talk about Atlas TI, I talk about other qualitative, I also talk about GPT 4.0, the new one, how to analyze your data with that. I talk about Julius AI and other tools that might be very useful for your research. So following me or subscribing to my channel will be also helpful.

[00:34:36] Speaker 1: And it's all for free guys. So capital like on this, everything is for free. Please, questions that you have, put them in the chat. If you have a topic that you would like for us to explore in the next live stream, please put it here, keep it going. So Dr. Adu, we spoke a bit on ethics and we have heard that some of these AI models are trained on biased data. So what else you can do to make sure when you're interacting with these AI models that there's diversity and inclusion and your data analysis, it doesn't come back so biased.

[00:35:13] Speaker 2: Yeah, I think that I always use this one as an example, right? We all have our own biases, right? We have your preconceived ideas and sometimes you are not conscious about. And sometimes you are also conscious about those bias. The same thing, remember the AI tools want to mimic our behavior. So they also have been fed with information that we have produced online and any place and we have our own biases that we have. So it's mimicking the biases that is in the environment or in our society. So you always have to draw the system attention. The same way when you are doing research, you do a little bit of personal reflection, making sure that your background and preconceived ideas will not overly influence how you are making sense of your data. The same thing, we call it bracketing. You can say that, okay, you can say that I'm aware that, you know, you can have a nice communication. I'm aware that you have been fed with a lot of information that may be biased. So for this analysis, can you reflect on a bias that you have and set them aside so that they will not overly influence what we are going to discuss? Letting the system be aware of it bias is very important, right? And then when it also depends on how you ask your questions, right? Are you asking in such a way that are you also asking a bias question, right? Because if you ask a leading question, it will also more likely to follow your lead, right? So you also have to be aware of, you know, asking not, you know, that balanced question, the question that is not leading to left or right. Or you can say that one example is if you say that, oh, can you go to the data and develop themes? Remember, it's developing recurring information. So this means that technically people who are minority voice will not be heard because it's giving you information that a lot of people are talking about. So you can prompt the system that, oh, can you review the themes and see whether there's something that you saw that is unique, but you did not notice. So you are letting the system do self-reflection and say, okay, yes, you give general information. A lot of people are talking about it, but a small number of participants are talking about something that's so unique that you didn't pay attention. Can you pay attention to that? Can you extract that information to me? You see how you are leading the discussion. You have to make the system be aware of its limitations and trying to also draw rich information from a system. So that's what I would do. It's not going to be perfect, but I think that questioning the results and also drawing the attention about the limitation that they have is very important.

[00:38:11] Speaker 1: And that's a very key takeaway. We just ask the system or the AI model to do self-reflection because it's capable. So we are right up until the very end. Please, if you have questions, keep them coming in. Earlier, Dr. Odu, you were trying to share with us a platform to compare different AI models. Now that you're able to share your screen, if you could walk us through that and then we close if there are no burning questions.

[00:38:43] Speaker 2: Yes. So let me share my screen now. Let me see whether I can find... Okay. Sometimes sharing is... Okay. Let me try one again and bring it here.

[00:39:04] Speaker 1: And in the meantime, guys, please remember, hit that follow button, like, subscribe, and share. And we will be having more of these live streams. So we'd like to hear from you. What topic would you like to have covered? Questions that you didn't think of now, put it in the chat so we can have a look at it afterwards. So we're trying now to bring up the platform that you can compare different AI models. So let's see if that works out today.

[00:39:33] Speaker 2: Okay. So let's see. Let me copy the... I think now you can see chat GPT, right? So let me copy and paste the link here. Okay. So this is the platform I'm talking about. It's called arena.lmsys. So when you go there, you'll be able to compare models, right? So as you can see here, first, you can even go to leaderboard. So leaderboard will tell you the model that is now popular. So you can see that chat GPT-O is, you know, has a higher score, followed by cloud 3.5 sonic, right? And then Gemini from Google, another Gemini GPT-4 table. So you see how, which model is like on top, right? Now you can also compare models by going to side by side, right? And then you can choose the model. So here you can choose, let's say, cloud 3.5, right? And then you can also choose here, maybe, let me, GPT, right? And then ask the system a question, right? So let's say, you can say, what is maybe qualitative analysis. And see how both are going to, you know, provide that information. So based on this, you'll be able to know, oh, I think that for questions such as this, I will use cloud AI, because it may be providing rich information compared to maybe chat GPT, right? So you see how I can compare and then decide, right? You can ask further question, and then you can compare. And then you can judge, right? Okay, I think you can say A is better or B is better, right? So this is software that I always use if I want to compare and see which model is good, right? Another software that you can explore to get, let's say, we call it free models, because they are, we call it open source, right? When you go to Hagen Face. HagenFace.co, and you go to Hagen Chat, you'll be able to, you know, explore some of the models that are open source, you know, the ones that are free to use all the time. So you can explore. Because of time, I cannot go and show you all the processes. But when you go to assistants, you see that people have created their own tools under that, and you can use them. And you can create your own one too. You can create your new assistant by clicking on new assistant. And then here, oh, I think it's gone. I don't know, it's vanished. Where did it go? Okay, let me go back again. So I go to create. And then when you go to model, you can choose the one that you want. You can choose any of the free models around. Most of the time, meta one is the one that I always choose, right? Llama3, that's the one that I choose. And then you give a little bit of instruction. Because of time, I cannot provide you all the information. But you can create your own AI tool and use it, right? Like the one that I've created here. When I go to my site, let me go here. I've created a lot of AI tools that you can also use. Let me share this one in the chat box. I don't know whether you'll be able to get it, but let me try and share here. Yeah, so yes, you can explore and see which one that is best for you. But if you need more information, you can contact me. I'll be happy to provide you all the support. And I also provide consultation too. If you want to learn more about any of the tools or you want to analyze your data, you can meet me. And I can help you with all this.

[00:44:01] Speaker 1: Okay, we're right on time. So thank you for tuning in. Remember, follow, like, subscribe. Check out Dr. Adu's YouTube channel. Lots of free resources there. Follow me if you're on LinkedIn. This is the quickest way, easiest way to be notified when we drop another free live stream. We'll be going through your comments, going through all the chats, and we'll definitely see you next time. Thank you. Take care.

[00:44:26] Speaker 2: Yeah, thank you for your time. And thank you, Anne. You did a wonderful job. Thank you. Okay.