EP 209: [Interview] Artificial Intelligence- How ChatGTP transforms Education with Dr. Mfon Akpan

Dr. Mfon Akpan Assistant Professor, Accounting Methodist University in Fayetteville, North Carolina.

This is his 3rd appearance on the podcast after Episode 92 and 194.

In this episode we talk about the Artificial Intelligence chatbot ChatGTP, how it works, the impact of it in education and more AI.

Listen to the episode on Apple Podcast, Spotify or other platforms.

Audio

Video

Clips

EP 209: [Interview] Artificial Intelligence- How ChatGTP transforms Education with Dr. Mfon Akpan

Dr. Mfon Akpan Assistant Professor, Accounting Methodist University in Fayetteville, North Carolina.

This is his 3rd appearance on the podcast after Episode 92 and 194.

In this episode we talk about the Artificial Intelligence chatbot ChatGTP, how it works, the impact of it in education and more AI.

Links to the guest:

Mfon on LinkedIn: https://www.linkedin.com/in/mfon-akpan-5702325/

Mfon's website: https://mfonakpan.com/
Mfon on youtube:https://www.youtube.com/user/MFONHELPSACCOUNTING

Give a review:

If you enjoy this podcast would you be so kind as to leave a short review on Apple Podcast/iTunes? It takes less than 1 minute and really makes a difference to convince future interesting guests to join me for an interview.

Connect with Jens Heitland:

Sign up for my weekly Newsletter ("Connect the Dots") where I share the innovative articles, books, podcasts etc. I discovered during the week. https://www.jensheitland.com/newsletter

Follow me on social media: https://linktr.ee/jensheitland

My businesses Businesses:

Heitland innovation: https://www.heitlandinnovation.com

SUCCEED Leadership Education Platform: https://www.wearesucceed.com

Applied Innovation Transformation: https://ai-fortuna.com/

Transcript:

(This Transcript is AI generated)

Mfon Akpan

[00:00:00] Hello, and welcome to the Human Innovation Podcast, the podcast for innovative leaders. I am your host, Jens Heitland, and today my guest is Dr. Mfon Akpan. 

[00:00:12] Dr. Mfon Akpan is the assistant professor of accounting at Methodist University in Carolina.

[00:00:20] This is his third appearance on the podcast after episode number 92 and 194. In today's episode, we talk about chat, G T P Chat. G T P is an artificial intelligence chat bot. And we look into how it works, the impact of it at education and more please welcome back to the show, Mfon Akpan.

[00:00:49] Hello Mfon, welcome back to the show. It's great to see you again. 

[00:00:54] Hey, it's great to be here Jens. Thank you so much. 

[00:00:57] Looking forward today, we have a special topic for everyone. So we are looking into chat, G T P. But before we do this, let me. Picture what chat G T P is and how the world got to chat. G T P.

[00:01:13] OpenAI is the company behind chat, G T P.

[00:01:16] And OpenAI is a research laboratory which has a nonprofit arm and a for-profit arm, and their goal is to develop friendly ai, which is artificial intelligence for humanity. They have done a couple of awesome project one, which is DALL-E, the deep learning, digital image production. Then they have had the first GTP-3 which is produce human-like text open AI five, which is playing video games. Then they released in November last year chat, gtp, which is a chat bot. They have an open AI codex, which is text two programming code, and a lot of interesting things coming for sure in the future.

[00:01:56] So chat, G T P is a prototype which was launched November 30 2022. It's now out in the world for a couple of months only, and it's valued already 29 billion US dollars, so huge thing that's going on, and it's it feels a little bit there. The first time. AI is really touches the mainstream on different usages, then behind the scenes so that we can use it in d.

[00:02:24] Mfon I'm really looking forward to explore what chat g t P means for education for us as humanity and so on. Let's go into the first topic. What is chat? G t P? Let's talk about the basic function of chat, G T P for those who have never heard about it it 

[00:02:42] is a chat bot. You can access it through a website.

[00:02:47] So you go to open ai, you click on the link chat, g P. and really you sign in and you can start asking it questions. It will give you answers. You can ask it to do different tasks. So you can ask it to write you memos, letters content. It can even solve problems for you as well. And when problems, in my case, accounting problems, you can put accounting problems in there and it'll give you answers as.

[00:03:18] So it's a very interesting tool. 

[00:03:22] Yeah. And I have used it for I was just playing around with it over the last three months from day one. I had, I think I got excess, I think it was still a waiting list in the early days. And I got excess after two days, and I was playing around with it until today, and I'm still doing it.

[00:03:39] So I've used it for those who are interested. I've used it for building summaries of wider text, which I have produced. But because I'm not native English, my text is not perfect. So I've used, hey, write this in a professional language style.

[00:03:53] Yes. And then it's writing it in a summary, which just sounds awesome. And then you can even write it in a humorous way, and then it's writing it in a humorous. It's fascinating what's already possible with chat G T P and it's still a prototype. It's not a real version yet because it is a limited of data set which can be reached with it.

[00:04:13] They're charging me for it. I signed up for the plus version already. Oh. , 

[00:04:16] that's 

[00:04:17] awesome. Yeah. I signed up for that prototype and you're right, you can use. Differently translate different languages or having to write texts in different languages for you as well. And I just thought about one thing you said about the valuation of 29 billion.

[00:04:33] I would argue that the company is worth more than that. Yeah, more than 29 billion. 

[00:04:40] It's already a couple of weeks ago, , so it's just 

[00:04:44] increasing. . Yeah. It's increasing. And you have to think about the, we were talking a little bit about this earlier, the execution. So yeah the way that they've executed the rollout of this to me is just amazing.

[00:04:59] So November 30th they open up the free website. I think it was five, it was five days. They got a million. Users. Yeah. Now they said active users. So I don't know what the, if that's Dow or, daily active users, but they got a million users. And I know that Netflix took, what, three and a half years to get to a million users.

[00:05:22] Facebook took a, I think about 10 months to get to a million users. So from thinking of it from that standpoint, , they've set a benchmark for anybody else who's gonna come in the space. Yeah. And from November 30th until now, so we're in February, who knows how many users they actually have. Huge.

[00:05:45] And then, yeah, it's huge. And that's the benchmark. So now whoever comes out, they've gotta think about, I've got to get to a million or million plus in five days, or I'm still. . So I think that's going to, that would create, they'd have to spend money potentially to do that. And as far as marketing and then you had me thinking about that cuz I just signed up for the chat, g p t plus.

[00:06:12] So their $20 their paywall, their premium version, and I was thinking about the numbers. . From what I know, so they've got a million users in five days, and if we only count that million users and they're able to convert optimistically 10%, so that's a hundred thousand. A hundred thousand times the $20 per month, yeah. $20 a month. You've got the $24 million. in a year already, so you can project that out. Yeah. I don't know what their conversion rate is. And that's just based off of, can you imagine that's based off of five days? 

[00:06:51] Yeah, five days. It's most likely. Way higher. . 

[00:06:54] Yeah, most likely. Yeah. So that's why I say that valuation, just thinking about it.

[00:06:59] And you know what, here's the other part of it. The, what they call the earned media. . So they've been get, we are on your podcast talking about it. And people are posting it, people are talking about it on TikTok and all of this word of mouth that they're getting is also fueling it and, with the Microsoft and the teams.

[00:07:22] So that's value too. Yeah. So whoever's coming in this space with something similar, is going to have to spend in, in my Unless, and even if they have something, a magical product, they're still going to have to spend people are gonna, they're gonna have to outspend on awareness, so I it's it's an interesting situation.

[00:07:43] Yeah. And it's quite interesting what I've heard. And now it's, I think, public as. That Microsoft was investing it in, into it, and they built a partnership early on, which means they're already looking into bringing it into the Microsoft ecosystem where you can basically, I guess in the future, do that as part of using Outlook, as part of using Microsoft work, Excel, whatever you do.

[00:08:09] If you just imagine that is just built into these things, it's incredible. What can. You're 

[00:08:16] absolutely right. 

[00:08:16] They just announced they're going to roll out a premium version of teams. Yeah. Powered by ChatGTP that will trans, so I know from the way I understand, it'll transcribe your teams meetings for you so you can go back and ask it, oh, hey, what did Jens say?

[00:08:36] Or what, what was said in this? Or you can translate it to another language. or if let's say we talk about in our meeting about an appointment, it it will write up an email and intuitively do things based off of the text in the meeting. So that's happening, I think it's $7 or $10 a month that they're gonna roll that out.

[00:09:00] And I, when I saw, I saw that, I said, oh my goodness, because you have to think about. The, so we were talking we were talking about a lot of stuff before this, but you think about the marketplace. Yeah. You think about now you're if you're using something like this that, now later on you can go back, Ooh, I didn't remember what time that meeting was.

[00:09:20] Oh I need to figure out Jens asked me to do something. Let me, now you can go back instantly and get it. That sets the. . standard So now you're, you are using, so either A, if you're at a company that's using this, you don't wanna go back to using Zoom or something that doesn't have it, right? No chance, the flip, no chance.

[00:09:43] And plus now you are more productive. Now you're able to get more done right and do more work in the same amount of time. So that adds value as far in the work. , you learning that, that having that technology and using it brings value to you as an employee. So it's there. There's many different shifts and dynamics that are going on.

[00:10:09] Yeah. 

[00:10:10] And if we just go very practical because. For, not everyone is into it yet, yet , but I guess after this podcast episode is led at least two people more. But if we just take your example, so you are doing a basic department meeting, which is on teams. You're recording the teams meeting where you have discussions on what our goals for the next couple of months.

[00:10:32] What are you doing? So normally what you have done in the old time, you. Someone who is documenting it and writing basically a write out, out up of the meetings. This are the meeting notes and that's documented. This are the task as part of the meeting notes and that needs to happen and that's what we agreed inside of the meeting.

[00:10:53] This is done as we speak. So this is this basically done through Che GTP straight away. and then what you can have, like you said, is not just having the written text of it, you, it's all searchable, so you can type in, Hey, what was exactly the date when we wanted to go out for dinner next summer? And then it comes back with the answer. , or you can most probably even program it to be able to do TA task management after that, directly integrating these things, if it's inside of the ecosystem of Microsoft into the task manager of the person that has got caught the task. We just imagine this, how much steps you normally would do and how much losses you have, and people not putting it into the task manager and not remembering or not reading the notes.

[00:11:42] , it's so much more. 

[00:11:45] Yes. So you see that, that productivity standpoint, I'll give an example of from an education standpoint. So if you think about online courses, and in many online courses, they have something called the discussion board, where there's a question that's being presented to the class online in the, in what they call a learning management system.

[00:12:09] And then you have to write so many words. And then, you do your initial post, let's say on a Wednesday, and then you have to respond to your classmates post at least, one or two times. And you have to have a certain amount of words to, to your initial post and to your replies. So what you can do right now and why educators were in such an uproar.

[00:12:33] You can as a student, you can go to chat G p T, which is free. Right now. And go copy your question from your discussion board in your class and put it in the little bar on chat, g p t. Click enter and it will answer that que And you can even tell it how many words you need. Exactly.

[00:12:53] Yeah. And it will answer that question for you. Then you copy and paste it and put it back into the class, into your cla and post it as your. and then take the replies from your classmates and you can do the same. So that's the big concern that on the side of educators with the technology.

[00:13:13] But it's, but at the same time it, it's a conundrum because especially if you're a university educator, cuz you're preparing students to go into the work. And now you know the, these are tools that are gonna be used in the workforce. Yeah, exactly. So you gotta prepare 'em to use it. But at the same time you're saying wait a minute, they can use this to really do most of their work, but at the same time we want 'em to use it cuz we gotta prepare them to compete.

[00:13:41] Then my other thing is, . If I was a student, I'd be using this because now I can finish all of my work faster. . Exactly. And go and do other things. You know it, it might've taken me an hour to do those discussion board posts, but now it takes me maybe seven minutes. . 

[00:13:59] Yeah. This takes us a little back to our first podcast episode number 92, where we talked as well of the changing landscape of education.

[00:14:09] And I think that's a key part of that. One of the things I believe, and if we look backwards, we have learned and that has changed now already, but when I was in school in the eighties and 90, . It was like, you need to learn and you need to just repeat what you have learned. It's not really connecting it to wealth perspectives, getting your own perspective and learning how to learn.

[00:14:35] It was just, okay, I will remember this for the exam and then I forget it the next day. So all of that is not needed anymore in a way because chat g d P can give you the. But I think what we still need as humans is to evaluate the answers we get. So have base knowledge of systems based knowledge of understandings.

[00:14:56] Maybe not the detailed knowledge of, hey, which day on second World War this happens. It is more about, Hey, what implications had the second World War and how important was that in history for things? . But I think that's a key valuable thing where education needs to relearn of how, what are the things we need to teach people.

[00:15:16] It's not about remembering specific dates, it's more the importance and the human aspects of that. 

[00:15:24] Yes, exactly. Understanding it. So I can give you an example. 

[00:15:28] When this came out I was asked to join a research project. . So it's 300 accounting professors getting together. So you can only imagine how this is working.

[00:15:37] , we're led by a Dr. David Wood, who's at Brigham Young University. He's really quarterbacking and putting all this together. And the thing we're doing, we're taking our exams, so our accounting exams assignments. , putting that in ChatGPT and seeing if it can answer the questions.

[00:16:02] So I used one of my final exams for this research. And the exam has the students, they get some information, they have to prepare financial statements, they have to prepare a statement of retained earnings and income statement A balance sheet and a statement of cash flows with the numbers that are given.

[00:16:27] So I put these, so now we started this in January, so this is key. So you gotta remember chat. G p t was released November 30th, so it's the last day of November. So then you got December. and then all of this was being, the data and everything is being put together. I, you know what, I'll take that back.

[00:16:48] Put together December and in January, so it was through December, through January. Now this where I put in my data was in, I wanna say January and ChatGPT the math. So it would do the, it, it would put in the. The listing of accounts. So the formatting, it did a good job, but it could not do the, it do the math.

[00:17:13]  So it did, it failed the exam, unfortunately. So with that being said, yes, the students so what I saw there at that time, so this is key. The dates are key. So at that time, I saw. . Okay. The students could use this cuz it's gonna format and put everything in the right place, but the math is off.

[00:17:34] So they need to understand what the, how the numbers work and how to put it together to check it so they could get done with this a lot faster. , right? So they can do the final exam pretty quick if they understand, like you said, are able to evaluate now what happened on January 30th. They updated chat G P T and it's better with.

[00:17:56] So I, I, when I signed up for the plus version, I just took one of the problems, from that exam that I used in for the final and for the research. I put that same problem in there and within literally 30 seconds it answered the pro it created retain earnings, incomes, it answered the whole thing and it answered it correctly there.

[00:18:18] And there was no, and what was interesting, there was no prompt, I didn't say answer the, I just. and pasted the question. Yeah. And it just, it followed the question what you were supposed to do in 30 seconds. It had it. Wow. Yeah. But, so I think the key is, again, now you have students that you can just use the AI to do the work, but at the same time, I think it's important to understand.

[00:18:47] So they still need to understand how it's put together. But again, going back to that efficiency, you're at the job. You had the you're given this task, you can finish it. In 30, it's done in 30 seconds, whereas it might take someone about 30 minutes to put thing everything together and check through it.

[00:19:06] This is where we are. 

[00:19:08] Yeah. The interesting part and that, that plays to, to, to what you said with the dates, it's learning as often as people are putting things in. And what we have seen is that the demand of usage is so high that it was breaking down in between. Yeah. I think specifically over the last couple of weeks

[00:19:26] If you just think about that as more, we use it as better it gets, that's how the algorithm works with ai. It and we are using it since November. . So imagine we, we think about a year, November this year. It's already that incredible end. Three months. If you now look forward, what's going to happen?

[00:19:47] It's huge. , it's incredible what this tool can do and or the algorithm, it's super powerful, which is as well, a little bit daunting in okay. What's how it can be dangerous in the wrong hands in the future. 

[00:20:02] How how do you think it'd be dangerous in the future? For, 

[00:20:05] for me it's more, it de depends, of course, on the data sources.

[00:20:09] . But if he can research everything like how to build an atomic bomb, which I think you can already on, on the internet today, but if you get more detailed instructions and everything, there are people who should not know how to do it, and they can. And I think that's just one of the examples or how to overcome.

[00:20:34] Estate in a different way. So how do you can do things and you will get answers that will work because of the data you can get. And that's just if you think about terrorist attacks and these things, I'm pretty sure you could get so much information study, you could do things way easier than today where you would need years to get all that information.

[00:20:53] So it's, it really depends on what the machine is feeding. and as well how it is then protecting. And that's for me the interesting part if we look at open AI and the promise of open AI of building something that's good for humans. So yeah it's interesting where this will go over time.

[00:21:11] I'm super excited in a positive way, so I'm not seeing it negative. Yeah, 

[00:21:15] I'm excited too. And that's a valid point. And. and I think with the different so that was a good point. When you think about these technologies when you mentioned that and you have to think about regulation or issues that may come up.

[00:21:30] One, when you thought about that, I thought about the metaverse. So there's uh, issues in the metaverse, which as far as I know, Altspace br. , which is Microsoft, again, . They've, they they've addressed it as far as assault in the Metaverse. Huh? So if you go into their platform, you can create a space bubble.

[00:21:56] So other avatars can't get too close to you , or come and, inter touch you or get, move in too close so you have your person. space. And these are things, you think about, wait a minute, it's not me, but it's an avatar but isn't me. But then someone is getting too close and and then when you get inside, so if you go into these platform and you're using it, it's spooky.

[00:22:21] Some of 'em are really sp, I don't mean that in a bad way, but people eyes are blinking and you see these characters and. Okay. You say yeah, it might, someone getting too close may impact someone as well. So those are issues. So you, with these technologies, those are things that you have to think about, reg, how to regulate and what ways to regulate it.

[00:22:42] Who should regulate it? Should it be the government, should it be self-regulated? Who comes up with these things? Then the other thing that, is that we are talking about chat, G B T, but what about the other platforms? Because there's other ones, and there's going to be other ones. I think Google is meeting next week to talk about their search engine and everyone's saying they're probably gonna, announce some sort of chat bot, 

[00:23:05] Platform or something they have it already ready because what if you, I think it was more than a year back, if not even two years, where they, I don't know if you have seen the video where they, there was the phone call making a hairdresser appointment that was totally done by ai. So it was working already at that time. Which, that was a specific use case, but I've heard that it is working. They just haven't released. because they said it's too dangerous for the world right now in the way they have done it.

[00:23:34] So I am pretty sure they have something up their sleeves. Not sure if they're that far comparable to Jet Chat G T P, but it will be an interesting one. . 

[00:23:42] Yeah. You gotta, th you have to think about it. It, if they got. a million active users in five days. Yeah. You, so I, and like I said, I was thinking about this.

[00:23:55] Imagine the, so this is what I was thinking about. They've got a million active users in five days. I, we don't know how many users they have. But now they've got a conversion base. They got me. Yeah. How did they get me? I signed up for the chat, for the free version. So they sent me an email, said, Hey, you can join in on the plus 20.

[00:24:20] I signed up. So just imagine if they, so if they start rolling out other features for example, what I was thinking would be really great, some other, some research colleagues of. , we were working on a academic presentation. And we said, Hey, it would be really neat to have chat G p t to write the abstract.

[00:24:44] Absolutely. Yeah, we get into a zoom, so as three of us, one of us on chat, G B t telling it to, to write, according to these specifica, it writes it. Then we say, change this and change that. Yeah. I said, what if they rolled out a share? . So now just like a Google Doc or something, you can have three or four people come in and you can work on something asking it questions.

[00:25:08] So all they have to do is roll these things out because they have the users. So I think we, we were saying earlier, whoever's gonna move into that same space, the it's like a the clock is, Because the, because they're gonna keep rolling out more and more products and growing more and more, and it's gonna be, it's I won't say it'll, in my opinion, it won't be harder.

[00:25:35] It's gonna cost more Yeah. To get into that space the longer you wait. 

[00:25:40] Yeah. If you just think about, okay, it's embedded into the Microsoft ecosystem, which a lot of companies use, so that's already. , if you think about that. And then Google is not part of that, but Google is building their own thing. So if these two companies, which are already some of the biggest companies on the planet and most successful with as well, the most data sets, I guess if we take data in general, Google is far ahead.

[00:26:10] If you look into what they know about search engines through the search everyone is using still. and on top of that, YouTube, which is also owned by Google, if you just think this and you feed the AI with all of that data and everything is searchable. Every, that's massive. It 

[00:26:34] is. But you gotta get people to use it.

[00:26:37] And now , you got, you've got open AI and the chat g p team, they're giving it away for. a million users in five days. Now they, they ca and I love what they did because the marketing, everybody thought it would be $45 a month and then they release it for $20 a month. Yeah. 

[00:26:55] if you have enough users, that's okay.

[00:26:58] Yeah, 

[00:26:59] I guess you're right. It's okay. So you got the $20 a month and now it is with Microsoft Teams and who knows what else they're going to roll it out in. . So now you gotta start. How are you going to compete? You know how you gotta spend money? It's gonna, it's gonna and I think it's, lemme ask you this question.

[00:27:19] Yes. You said that they didn't wanna release it because it's too dangerous. So then why do you think they wanna release it now? Is it's no longer dangerous? No, I think they, 

[00:27:29] they will because they have the pressure. If, for me, it's more about the specific use case. Because right now chat, G T P. A lot of users are just playing with it.

[00:27:39] They're like, I do, I try to produce social media posts, , which work fantastically sales letters as an example. Hey, write this type and this style into an email sales letter that's working perfectly. So I think these are all specific use cases. Then it's more about if someone comes up with a very interesting use case that is,

[00:28:04] So just what Microsoft already said, they will include it into the Bing search engine, which is owned by Microsoft. That's a huge thing. But Bing is not used by everyone. Most people are still using Google for search. If we take worldwide, so if Google would build their AI engine.

[00:28:24] The same way they can just saying, okay, we, we can do the same what Chat GTP is doing. You can do this from now on, like next week inside of Google Chrome while you search. I 

[00:28:38] see it. I see it. But you got, so right now the open ai, the chat G P T, they got the front end and the back end. So they got the front end with the.

[00:28:48] So the search engine, the backend, they've got the Microsoft cloud. They're using that. Yeah, I see that. But at the end of the day, we, so we know chat, G p t, they've got the numbers. So they got the people on board. They got the people using, they need to convert or get people to start using it through being, so that one I see the question is, , how would Google integrate it into their, they've gotta come up with something and it's like they've gotta get people either from Google to go to their product or how am they gonna 

[00:29:27] integrate?

[00:29:27] why, That's why I'm saying they should integrate it straight away into the Google ecosystem, that it's not something extra. So you use Gmail, for example. It should be integrated. You. Google search, it should be integrated so that it's natively part of what you do and you don't need to go somewhere else.

[00:29:44] And I think that's a competitive advantage they have because they have way more than if we take total users of Google is still way more than what Chati g p has. So that's their advantage 

[00:29:56] right now. But here's the cost. It's gonna impact their ad revenue. 

[00:30:02] Absolutely. I think that's the key part.

[00:30:04] How do they earn money? . . 

[00:30:07] Yeah. And they gotta make money. So again, it's gonna cost them a lot. And it's it, I, it's just interesting because they, like you said that, and I, like you said, pressure because at the end it's one of those to, in just assessing it, I don't know the ins and outs of Google or open ai.

[00:30:27] but j just as you mentioned, you can see there's obviously some sort of pressure there and you can see the more they're gaining that momentum, the greater that pressure is because they know eventually and this is what's great about technology chat, G P T's not gonna go backwards. Yeah. They're gonna keep going forwards and getting better and better, and they've already got another version.

[00:30:52] if they keep rolling these. And so think about it, they came out with it November 30th. So you say November 30th, to December, and then January 30th, they get an update. Yeah. Does that mean they're gonna update it every month, and roll out new And we, we don't know. But they're sending that signal that they're moving forward with it.

[00:31:14] They're integrating might, so they're scaling, they're looking to grow. 

[00:31:19] and if you then think from the totality, if they open, I'm even not sure. I think they have open APIs, which other products can use the data as well. One example, I'm cutting all my videos, everything podcast wise is cut through ai. So I'm using AI and cutting my podcast, which is already phenomenally faster than I doing this like manually.

[00:31:45] Yes. And it's super du. Just thinking about adding that on top of that. Imagine. 

[00:31:53] Yeah. But you've already used it. It is a G P t three pro product. Where it 

[00:31:58] opening? No, it's a separate product. It's a different company. Yeah. Gotcha. It's called descript. I can put the link into the show notes for everyone who is interested.

[00:32:06] It's phenomenal. It's really good. . 

[00:32:10] Yeah. So this is how it's making things. . Exactly. It's making things v very efficient. And like I said I really I think open AI is going to, as far as execution and rollout, I think they're gonna be in the books for Yeah. they are. For what they're doing right now, and the way that they're executing.

[00:32:32] And growing so rapidly. And you're right, not everyone is using it, in, in that, at this point. Now, I was in a meeting and one of my colleagues was saying, I think I'll just, I won't give the percentage, but we were talking about it in a, in one of our meetings, and he was saying a large percentage of graduate students are using it to write.

[00:32:56] Yeah. 

[00:32:56] That's cool. You though, I have to say, correct me if I'm wrong, it's only in English right now, right? I have tried it in German. It didn't work. Uhhuh. , 

[00:33:09] okay. I thought you can use it to translate. Yeah. 

[00:33:12] May, maybe you can do, translate this to this language, but just having the normal chat version of it, I'm not sure if it works in different language.

[00:33:22] Gotcha, gotcha. But I'm not a hundred percent sure. Maybe that has changed already since I tried it last time. 

[00:33:28] Gotcha. Oh, so as far as people using it for papers and things, so just in English, 

[00:33:34] if they're Yeah that's why I was thinking because, and I'm saying that because if you take my ecosystem, my family and friends who are living still in Germany, they have never heard about it.

[00:33:46] So it's still early innovators. Now it's getting a bit more into the curve where it's like the early adapters are on it, but it's, I guess it's not mainstream yet. So it's not on that curve where you say, everyone knows about this. Everyone has heard about it. Just the people who are interested in tech, who are interested in technology develop.

[00:34:08] who are innovators, like you and I, we know about it. We have heard about the first versions of it, but still not everyone you talk to has heard about it. At least not in the, not in English speaking world. So there's 

[00:34:21] the pressure because eventually they're gonna roll it out in other languages. Yeah. And so there's the pressure to get in there.

[00:34:27] So you're absolutely right. About that, it. This makes me think back to when I worked in finance. So I worked in finance and I had a client he was a, he was from Greece, so he had, he came to Chicago he opened a Greek restaurant. He opened a Greek re. And, we were talking and I, he, he would talk about his restaurant and he said, my restaurant from day one it was successful.

[00:35:00] So I, I was just curious because there's we've got quite a few Greek restaurants, but this restaurant, he said it was more of a, a fast food. So you go in, you buy these, not really a sit down restaurant. You buy these I guess you call it Giros. Giros. . You buy the, you buy that. That was his main thing he was selling.

[00:35:20] So I was curious. I said, how is it successful from day one? And he said, and I come from a small place, Inre. I didn't go to school. Is it to any big schools or anything? But I know. and word of mouth. So he said what he did was on the first day he opened up, he gave away about $10,000 worth of food of heroes.

[00:35:44] He gi he gave it away. He said there was a line, like outside of the restaurant, people coming to get food. Yeah. And he said every day at lunchtime, there was always a line until he sold the re. And he said after I didn't do any advertising. I didn't do anything else after that. I just gave away the food and that was it.

[00:36:05] And then I was able to sell it. And he sold that restaurant more than what he had heard other people had sold their restaurant because he always had, people, always had a line. Everyone was talking about it. So when I see how they're doing this, how they're executing on. , I think about that restaurant, right?

[00:36:24] So if you had to, if you were, you come into a neighborhood and you wanna buy a, a restaurant, a Greek restaurant or a restaurant, and you see this one and they've got lines and you're watching it, and every day for a month they've got lines and lines of people in there.

[00:36:40] And then you've got another one that opened, maybe three months later. But they've got good food, but they don't have those lines. Yeah. No one's talking about it. You want to go into that one? So a hundred percent . So this makes me think about that that convers. . 

[00:36:58] Yeah. Before we finish up, let's look into the future.

[00:37:02] So if we record an episode end of this year, end of 2023, where chat, G T P and the development of chat, G T P is one year old, where is it? What do you think? What impact do we see in the world? And maybe if you want, we can talk about the educational world where you work in, and the implications of chat G t.

[00:37:26] In that space. 

[00:37:28] Y yeah, so I think we're going to see deeper penetration. So wider use. We're also going to, unfortunately, my thought is there may be a rift, so ultimately there's gonna be a, some sort of divide. And what I mean by that is it's gonna get better and better, but there's going to be more pay walls.

[00:37:54] and I think you're gonna see that there's gonna be more and more paywalls. And depending on the pri, the price points might increase. So you'll see a rift where you may have some people who are able to afford these tools, and then you'll see people who are not able to afford them. And then what I think in a year from now, you'll see more there.

[00:38:19] There may be more of a divergence. where we'll start to see those who have all of these tools and the latest AI are going to be the most productive, going to be able to move the fastest, especially in business, are able gonna do more, serve their clients better, move faster, versus those who are not.

[00:38:38] And then those who, cuz you mentioned early adopters, those who have not stepped into it or not even or learned about. Are gonna be catching up. So I, I see that, but I see definitely there's gonna be more and more penetration, more companies coming up in different spaces. So we talked about cactus ai.

[00:39:01] So they're in the education space in different areas outside of G P T three. , there'll be other companies coming in there as well. And you're using one, so this the software you're using to edit. So you'll see more and more of that. Yeah, agree. 

[00:39:18] And I think it's also if we look ahead, I'm even predicting asset as it will be that some companies will not survive that because they're built on helping companies to be more efficient.

[00:39:33] Hey, we outsource this and we put the human hours into that for you. And this can be done like in 30 seconds, in two minutes, or in six minutes rather than in an hour, and it's done by a tool rather than a human. So I think that's what we might see as well, because some hours will not be needed anymore.

[00:39:53] So companies need to either find new ways for allocating these hours. . Other interesting ways, or worst case, they need to get rid of the people and or they're, as a company, as a business model. If they're not innovating themselves, they're not needed anymore. 

[00:40:11] I agree. I agree. I think there's gonna be a squeeze.

[00:40:13] Yeah. So you're gonna see these, if they're offshoring and they're having people to do tasks manually, I think those companies will start using AI to costs and mar and have fewer people do it eventually until the technology's gonna get so good that yeah they'll say, okay, we can do this with AI here,

[00:40:38] We can buy the program and use it here. But yes, absolutely, there'll be that squeeze eventually. But I think in those, the offshoring, they'll use it and to cut. And because you made a good point. There's early adopters, so there's going to be a curve. So once more and more people start using then I think things will shift in that other direct.

[00:41:00] So that's absolutely good. Good point. That's a good 

[00:41:02] prediction. Yeah. Let's see where this will go. M fun was a pleasure to have you again on the podcast. I'm looking forward to record an episode end of the. On this topic again, with the and of course we might have a couple of other episodes in between

[00:41:19] That'd be great. It'd be my pleasure. My pleasure. Yeah. Thank you for all you do. This is an amazing platform. 

[00:41:25] Thank you very much for joining. What's really a pleasure to have you again and looking forward to stay in contact as we do in WhatsApp and all the other tools. Thank you very much.