Hello, today I came back to meet with the sequel of ChatGPT. This time it's strange, so I'm forbidden. Do not use ChatGPT. If you don't know this, tell me first. So this talks about the truth of ChatGPT . Let's see what truth we as users Whether it's a marketing teacher Marketers, students, researchers, anyone who uses ChatGPT, which is now so prevalent. It is necessary to know this fact first. Before you start using it, Aha, OK, so today we're going to talk about it. Don't use ChatGPT if you don't know this with Ajarn. Before going there as before, please press like, press share, press subscribe, and most importantly, ring the bell so that you don't miss it. When there is a new clip, and what clip would you like to hear about? Comment to tell Teacher Ne under the chat a little bit. Okay, don't use ChatGPT if you don't know the truth of this. Today, I will tell you about it. The truth that we must know before thinking of starting to use ChatGPT, the first question is, how modern is ChatGPT? Of course, ChatGPT will enter the TARDIS in all digital databases in a huge way. Some people call it Metabase . However, it's good. It has been revealed It can only be reached in September 2021, that is, September 2021 only . It can't be found yet, well, so you need to know this one. If you ask anything that is very updated It won't be able to answer. and including if asked anything It can only answer according to its Metabase until the year 2021. This is something you should know. Nee has an example for you to see. This one, Nee went to ask ChatGPT, just asked fresh and hot to make this series of videos. Go ask about gossip. ChatGPT You tell me about Lisa Blackpink that there is news that she is dating with LVMH. Actually, write as short as this. It can be answered. I can't answer, I can't, please tell me, sorry, I only have knowledge in September 2021, nothing, know anything about this. And then sit down and tell the history. Lisa Manoban, it's a K-pop group from Blackpink that, uh, went into YG Entertainment. What's that and that, and LVMH is the group that makes louis vuitton, I can only tell you this, but I say that I don't know how to involt because my knowledge is only September 2021, sweet confession. very much, so everyone Don't ask for anything that's updated, you can't answer until ChatGPT has a new update. Must continue to follow up whether it can be updated more or not In addition, the second fact What data sources does ChatGPT use? The answer is simple. I can't tell you. It's not like Google Bard. Or uh, uh, it's not like Bard . Search Engines, so ChatGPT, please don't tell me, I'll have an example for you to see . Regarding ethics, tell me what ethics consists of. But where does the source come from? I don't know. We have to search for it ourselves. If you want to know the principle of autonomy, in case you believe him too. We have to find out for ourselves the principle of autonomy who said this, the principle of beneficience who said this. Therefore, I do not tell the source of information that I have searched that all the principles that have been written here. Where did you search? Well, let's continue. How accurate is it? This must be a question that many people, uh, must be thinking in their hearts, of course, how accurate is what I got. Like I said, ChatGPT makes you think like a guru. I thought that they were our friends. who wear thick glasses all day, only search on a computer, something like that, so think about that we have a friend who is an encyclopedia guru, who sits down to search for information every day to answer us, so sometimes if We rush to find answers. I can't answer. He processed the information he had. and then guessed in order to answer Anotai There will be around 20% guesses. This is a number from the Australian Library Network . Guessing numbers. There will be about one in five that will be randomly guessed data, make up the story for you. Don't forget, ChatGPT is a Human Language Processing, so it will be able to generate messages. all the time This is the capabilities of ChatGPT. Yes, this is an example. I give ChatGPT to say, uh, you can help me with a literary review, that is, help me write a review of literature related to uh, marketing mix factors. Please give me some organic vegetables. Of course, oh wow, it's something that's very specialized. It's an academic thing. Second, it's only focusing on organic vegetables , so it's very narrow, and in a short time, two or three minutes, not even, generates for about a minute, so ChatGPT is called Use random guesses. Random guess to write this story for me. Write a review of mana for me. Which Nee used to say when talking about ChatGPT in research Because Nene stipulated that it be put in terms of references, please come along, so he definitely can't do it in terms of references. Because references, you have to go to, uh, Google Scholar. You have to go to, uh, different journal databases. ScienceDirect What kind of stuff is this? PubMed, damn it, it's hard to get into, and ChatGPT can't get into it, so what you can do is generate it. I don't know what What is this Hunter? Bennett and Preston, 2017 analysis of customer preference for organic food like this. A wallpaper like this, look it up on Google Scholar and it's not real . It doesn't really exist, so this must be Be careful. that it is in the speculative part of ChatGPT, that is, what is very specific, what is in the database that it is closed, it is very specific, usually you have to be a member to access or if you can enter General people can enter, they can enter in a way that they can't see everything like that. And then there must be a professional processing, a bit of professionalism. Well, ChatGPT will choose to generate a Human language straight away, where, well, here it will be some fabricated story. So, so, the review regulator's side in this research is composed of the whole Krabi. Okay, come to ChatGPT's response. Will ChatGPT answer our question well? I have to tell you. Depending on the prompt, the person who asks the question must be prompted to be prompted well. You may be prompted too. You can see the example of prompting. There is a software that is given. It's a plugin called AIPRM, so you can use it. I will leave the video card for those who are interested to use it . Write it concisely, clearly, use simple but clear words, don't be ambiguous, don't use too ambiguous words, and don't do it in Thai. because Thai language Is training Thai language not yet is not yet strong, use the prompt in English But if you want an answer in Thai language, that's still enough. But the prompt must be in English. It's mandatory. And tell me that the answer is in Thai language. Dammit. They can answer. Okay . I want to know about the prompt. There is a video already. Is it neutral? ChatGPT is neutral ? Have an analysis of the information for us before answering. Both the information that is, well, the left side and the right side. Analyze it properly. I can tell you. There is no such analysis, it is not impartial. ChatGPT's bias depends on the data that ChatGPT enters, which data it enters, and which data source it generates from. He didn't tell either. Therefore, we cannot detect bias at all. You can only look at it, naha, only see how we compare with what we analyze ourselves. And this one is Bias or not, so there is definitely no bias about neutrality therefore, must be used with caution. Especially the article that is analytical, order it to analy something, you have to compare the analy yourself as well. Just use their language structure. Ah, come to the last one. Lastly, is ChatGPT honest with us? That is, he protects the secret. Protect our personal data, right ? What is recommended? No, why? Because ChatGPT is caused by Getting everyone together. Enter input. The more users we have, the more. Let's join input. ChatGPT only gets smarter, so ChatGPT will take the information that we input and use it for processing. Which may be processed and then able to reply to another user or cause our information to leak So this thing has to be careful. As I said, it's from Open AI company. The name says it's Open, so it's open experiment for everyone to try it out. And everyone teaches ChatGPT to be smarter, so love ChatGPT, use ChatGPT, but don't get lost until it's all gone. Microsoft and google are ChatGPT's competitors, let employees use it. But what is forbidden? Do not include company information. Don't give information called sensitive data because it will leak them. It's his ban. So this is a reminder to us and the facts related to ChatGPT, which Ne has provided. I hope it's useful. And if it happens, well, if you are interested in anything more, you can contact me, if you have comments, you have questions, you can send them under this clip. Then let's meet again. If there is a clip about marketing Don't forget to ring the bell. You won't miss it.
