conversational AI

DRUID Talks Ep #5 How Will ChatGPT Impact the Future of Businesses? with Tom Allen

DRUID Talks Podcast Ep. 5 unfolds the origins of ChatGPT, its applicability to the enterprise world, and the risks and benefits of such systems to society.

Episode #5 of the DRUID Talks webcast features Tom Allen, the founder of The AI Journal and an AI enthusiast, and Subject Matter Expert Kieran Gilmurray. See the full episode and transcript below.

Subscribe for the next episode!

Kieran Gilmurray: Hello, and welcome to a new episode of DRUID Talks. Today's episode is “How Will ChatGPT Impact the Future of Businesses?” And today, I have a very good friend of mine, Tom Allen, who’s the founder of The AI Journal and an AI enthusiast coming to answer all of the questions that you and I need
to know over the next period of time. Tom, welcome to DRUID Talks! How are you?

Tom Allen: Excited to be here, Kieran; I really appreciate having me on. I've seen a few of these and it's exciting to be speaking with you as we've done in the past and talk about some really fascinating topics. When you mentioned it to me, I wanted to jump on and get into the conversation with you about it. So thank you for having me on.

Kieran Gilmurray: Pleasure, sir! So right, let's jump straight into this and grab some of these answers. So, we'll get into how ChatGPT will impact the future of business.

So first question, Tom. I'm sure you've tested it, I've tested it, and I think millions of other people have done the same thing in the first week after its launch, but why is there so much hype around ChatGPT?

Tom Allen: I think there’s too much hype around ChatGPT as a starting point. I mean, there's a lot of
people talking around it, and from the research I've done, I mean it's not in the way it’s scaled and the way it's being brought in and the way it's been marketed, I guess to a certain degree - it's not massively revealing the way that it works, but I think it's because people have opened up their doors to, I'd say, Creativity - which is a big point of it. I mean, I've seen so many examples of creativity and business use cases and ways of it being implemented even to the consumer from people using it to pass university
exams. Right? For people to understand different language models and different contexts, a language, especially if you're an NLP, if you're in the NLP space.  So it's, it's a really great way.

And I mean, when I look at it, it's, it's something that's clearly caught the buzz, and I guess I don't actually have an answer for why it scaled so quickly! I've just watched the search results, and I saw some stats showing how long it took Facebook to get to like a million users, and it was like five weeks, and it showed Instagram, and it showed ChatGPT, and it was like a day or something ridiculous. It was just how it's come through so quickly… is remarkable! But it's a very powerful tool and definitely not going away. So I'm very, very excited to see what we can do with it in a business case, in a personal day-to-day case.

Kieran Gilmurray: It was interesting; it was fascinating for me to work out why did it suddenly explode?Because it's a good tool, but just ChatGPT 1 and 2 have been around for a while, and Google have done their own thing as well. So again, but maybe it just caught the Zeitgeist; maybe it was a moment sitting there when there was nothing else to talk about. Or maybe people quickly realized the power of it because it does seem powerful. But if I ask you this…

Tom Allen: It's like sliced bread. Sliced bread was around 20 years before it became popular. I always go back to that analogy. It's just like no one liked it, then they did a certain marketing trick, and now everyone has sliced bread. So maybe it's the same sort of deal that someone just found the right way to market to the right person. And that's how we are in a situation where it's going so big today.

Kieran Gilmurray: So, Tom, for those who don't know what it is, just give us a little bit of a definition, then. So, what is ChatGPT, and how does it work, please?

Tom Allen: Well, in my definition or the way that I look at it, it's something that's conversational AI that's run on a large subset of data, reinforcement learning and looking… I think the word is “derivative”. So it looks at very past history, which is why it gets its – obviously - challenges. I think we'll look at that a bit later. But it's really just a very well-optimized chatbot that can scour and look through and go and slice right through God knows how much data to get you an answer that's creative, reasonably actionable, reasonably helpful, but also has quite a lot of challenges on the F-score bias side, the reason why I think it's got a long way to go.

And I think you might have seen, Kieran, the update around LaMDA… getting absolutely criticized in the press. I think it was towards the end of last week, mid-last week, when they just showed faults, and they showed it was less than 6% accurate or something. And asked people to double-check that. But it's real, it's a really powerful tool for businesses, and it's something that… it's just a very sophisticated chatbot
that can use a lot of data, use a lot of machine learning and use a lot of reinforced learning to get you an answer a lot quicker, and a lot more practical solution, and in a lot faster way.

I had to tell someone what my definition was or how it works the other week. So, it's always a good question. It's similar to the question I used to get, which was, "Tell me what AI is." And as I walk it, I have 1001 different definitions for it. It depends on the context you put it. Or I've heard a thousand or more different definitions. But yeah, a very powerful chatbot, in my opinion, to keep it simple.

Kieran Gilmurray: Because it does go back, I suppose. You mentioned earlier on, you know, the variety of uses because I had an interesting one. I was testing it the other week, and I looked at a Word list, and I filled in the Word list, and I typed it into ChatGPT: what I want you to do is turn the list into a paragraph. And it did, you know, really nicely written. And then I said: now turn that into a paragraph as McKinsey would write. Then I used it as a search engine as well to look up data on the series that I was doing for LinkedIn,
and it was able to do that as well. So I'm sure that's why people are confused as to what it is or what it isn't.

It just appears to be a super cognitive tool that allows you to do 1001 things if you know what to ask it to do. And maybe therein is the skill.

Tom Allen: But I said it's the versatility, right? You can use it for fine coding, and you can go and do SQL checking… things that wouldn’t appeal to me. But people love that kind of area, but things that would appeal to me, like you can write code, for example… But you know as well as me, Kieran, that we both love to write. I mean, you've got your amazing book out, and sometimes you get the writer's block. I mean, this just takes that straight away! Put in a title, put in an idea, get… want to write a poem… For me, it's Valentine's Day today… If you want to write a poem for your loved ones, I mean, there are a thousand
more different examples. And they're really… that's what I say. It can’t be unbiased, but it really depends on the context you're using it. And if you're a… what we might use it for The AI Journal, if we've got an editor thinking of what’s the story or what's going to be a hit or a good marketer, it's, it's a great tool in that aspect. But again, it can also be used to help your coding to help you write a new song that's never been heard before.

So I can collect data and say, this is a combination that's not been used. I mean, in a business sense… I mean, you can see how important it is or how important it can be. I wouldn't say immediately, but over the next five, ten years, I always have the longer view for things like insurance, medical records, things
where it's going to be very hard to go through that data. And I mean, one thing I noticed with that,
or I'm very aware of what people that are experts in the field that run the conversational companies and similar to what DRUID is doing, is that the context can change everything. It's like if you say two words,
“Jonny bit the dog” or “The dog bit Jonny” it's the same words, but has completely different meanings
and completely different contexts. And I think that's, that's why it has scaled so much because you can use it for one area of marketing or you can use it for another area of marketing.

You could use it for creative strategies, or you could use it for creative writing, or you could use it for writing a certain piece of code or finding a bug in the code. I mean, there're various different angles… is how I'd look at it. And that's why I say in my definition, a very sophisticated, intelligent chatbot that you can get for free.

Kieran Gilmurray: But I suppose then that leads me to the next question. Just... Is there a shade of originality or creativity about ChatGPT? Or is it “pastiche” in that if it's been trained on historical data, cannot write a poem, cannot write a song, for example, you know, or is it just going to regurgitate and reconstruct and reorder what's already created?

Tom Allen: Aaah, that's a great question! It's… it's a tricky one to answer. I wouldn't… I don't think it's ever… I think the only way you can be original is if you come up with something yourself. I mean, that's something that I always try to stick to; you can use it… And I mean, most things in this world now, I don't think... it's very rare that you see something that's brand new. I think the last time I really saw something of significance or a top of mind. I mean, maybe that's something since then, but the iPhone just changed everything. Right? And I mean, it's a substandard one to use. But if you look back at it and you just think, if you had told me in 2000 or 2000, even the year before it came out, 2007, what, 2006, that you can have a web browser, music, all these things in one phone and it's going to be touchscreen, I'd say you’re full of rubbish! It's not possible, and it's just that now we take it for granted. We have God knows how much now… No one buys digital cameras unless you're a proper photographer and videographer, and maybe it will be the same that it changes trends, but I don't think it will.

Well, it will change trends… But I don't think just using consistent backlogged data and using machine learning... and I think it will lead to innovations, but I don't think itself will change much if that’s a good answer. Because I just don't see how it can. I mean, it's always going to be working on past data. It's never going to be... it's never… it can predict, but I can't see it creating, which is a hard thing to explain. I can't see it pushing off a production line, a brand-new product. It might give you a free drawing of an idea, but at the end of the day, it's still humans that make that decision. It's still I always say I'm never going to see a chatbot that sells to another chatbot. Maybe I will one day. But even then, you know, it's a business behind each business. So, you know, it's a person behind that business. So essentially, when you go to every cause, it's not in... it might be a chatbot selling to a chatbot, but it's still, you know... I've never seen a chatbot take the money and buy a mortgage, put it that way. But everything, it can be creative.

Kieran Gilmurray: Yeah, well, it's interesting. You could maybe have your digital assistant go and find the best mortgage based on that financial history that it was able to link up. But I get what you're saying there on creativity. I suppose we always compare it to humans, and it reminds me of the movie “I, Robot” when Will Smith said to the robot. You know, "Can you create a symphony from scratch? Can you draw like Michaelangelo?" And it was quite funny as the robot then turned right and said, “Can you?” You know… it's a little bit, you know, we're comparing crazy things.

I think there are only 14 cords of music; in the English alphabet, there are only 26 letters. You know, a wheel is a wheel. But it's interesting how we, as humans, suddenly create differences even in that mix. But I'm forever surprised, which is supposed why we're in technology space.

So, suppose I move on to the next question, what do you think makes it different from previous innovations in conversational AI?

Tom Allen: Well, I think it's, I'll say, the way it's been marketed is a big one because people understand it a lot more. And I mean, you see diversity in using it before, really I mean, that's something I get really, I guess… and you know my not anger but kind of annoyance with it is that people that aren't in work technically or don't really know unless you're doing a master's degree or something, what's available to you. And I think this has really opened that conversation about what AI is, and I've had all sorts of people that never looked at what I'm doing, like some nerd or some geek or whatever. He's involved in every tech,
whatever label I put. And now we're talking about it while we have a beer. And I'm just like... it's very thrown into our social lives very, very quickly and very... it's introduced a lot of people that necessarily wouldn't
know what AI is, and I think it sparked a lot more conversations about what it is and what it can do.

And it's kind of very similar to, I guess, what some people might say about big effects. Like… you're ignoring if you're taking climate change, if you're ignoring it and you're hoping it goes away, whereas with AI ignoring it and thinking it wouldn't take my jobs. And that's making it very real. And I'm not for or against that argument. I think it's a bit of a silly argument because it's going to take jobs. But I mean, time and time again, the research shows it is creating so many more jobs than it's getting rid of. It's just increasing demand for higher creative jobs. It's great since it creates other jobs, but I mean, that's a different conversation. And it always sparks a nerve because I'm like, what if you did a research? It's creating a lot more jobs that we don't even have the jobs for.

So it's an ongoing debate. But I mean, that's where I'd see it really as being a lot different because it's just... it's opened the floor to a lot more people that wouldn't understand it. I mean, the big thing is you don't know, I don't know many tools where I can go in, where people want to actively go in and try and use it. I don't see people wanting to go to an RPA tool, to an NLP platform or go to a piece of automation and start trialling it with this and go ask what film or what movie or something very, very personal to me. And once you have a hands-on experience, it gives you an “aw, this is quite interesting!” So it's a much more hands-on approach compared to how we’ve been using AI before.

Kieran Gilmurray: If, if I was to look at it in my personal opinion, maybe that's a bit like the iPhone. Is that because the interface is so simple that people are willing to use it a lot more, and maybe that's going to software design. But if you look back at the history of conversational AI, then what do you say is the difference between a simple chatbot, ChatGPT, and an enterprise conversational AI assistant?

Tom Allen:
So, I mean, I think you told me about it actually … the future of looking at everyone, every single person in work… And I think it's maybe part of UiPath’s mission to have a robot for every single person. And I think that's where it's looking at that question. It's much more applicable to businesses because I mean, I'd love
a personal assistant to work and being able to book things and being able to do things for us. But I mean,
what the chatbots before, you know as well as I do sometimes I don't even or at least I wouldn’t say, you know, as well as I do. But you can probably understand my stress when I spoke to chatbots before in the past five years I still think at times, you know, helping… I want to get through to an advisor whereas ChatGPT is just working on, from what I can see, on much more relevant data and insight and things that aren't just preprogrammed It's just using a wealth of knowledge from a library, God knows! I don't think
even if you put all 6 billion people on looking at that library data and be able to go through it all and in the way that ChatGPT can, or whatever else is coming out with Bard and LaMDA, all these kind of cool areas
that are going to occur. So what we do in search, But not just in search, as you said, in a business context.
I was speaking to a company yesterday with how they're doing it for medical records and insurance records.
And it's such a… it's such a big way for medical records and insurance records. And I think it's such a big issue around things like language translation. If you're in travel and if you're booking travel for 100 employees or whatever to get over that, or a thousand or getting over… simplification or dialog, it's so powerful and it's using that in a business sense. You can use it in a strategy sense. You're looking at your OKRs or KPIs of a business having that kind of assistant, it can get rid of almost what I call a writer's block or a strategic block,  which a preprogrammed chatbot that you're trying to book a certain or buy a certain e-commerce product through, or buy a certain plane ticket just can't do. And now I think it's opened up the doors to all of that kind of things to get very personal at scale.

Kieran Gilmurray: Now, it would be interesting, wouldn't it? Because imagine if you had a conversational AI system just watch and monitor what you're doing every day. I think of it: watch me. I would kind of hate if it described that as intelligence and started to repeat it. But if it could do stuff, you know, when I'm not there 24/7 like go and book the flight, sort out the calendar, sort out the meetings, you know, answer questions, complete a piece of research.

But where do you see ChatGPT as applicable to the enterprise? You know, what are those good use cases
that I'm describing or do you think ChatGPT will be applicable to the enterprise now or not yet, or ever?

Tom Allen: Now, yeah! It's 100% applicable. I just think it's got a lot of challenges to get over. I mean, there's a lot of things or areas that I just still, I mean the big thing for me is trust. I mean you've seen it all around fake news and I watched literally… I think it's Steve that posted it on LinkedIn and it shows you Leonardo DiCaprio doing a speech at the United Nations around climate change. And they've fixed it to have like six different people using his language or putting their voices over what you say. And they use like Kim Kardashian, the president, and all these different people. And I think that's a big thing for us. It's trust. It's making sure we’re... And I think a lot of companies… And I see DRUID do it… and it's the power of educational marketing so important that educational marketing can only work if you're being honest if you're being factual. And I think a lot of companies are switching towards that route because they know that helping a customer is what's going to get them as a customer.

And I think that's where ChatGPT can be very powerful in the enterprise… without diverting too much of the original question. But also there's a lot of things that one needs to be careful of. I mean, from the creative
side, it's brilliant. But in short, yes, it’s definitely for enterprises, it’s not going away. And if you think people are in powerful positions that have the spend, they're going to just be ignoring this when assessing what are opportunities it unlocks... That’s just not going to happen! Well, maybe it will. I can't see the future, but
I just don't see it happening that way.

Kieran Gilmurray: Yeah, it's interesting because I sort of look at enterprise use, and I do worry about people's jobs… It's a bit like, you know, and without getting too excited about ChatGPT, just looking at generative AI… someone gave the comparison that, you know, 100 years ago tractors took, you know, thousands if not millions of people out of the fields and put them into factories.

Tom Allen: Yeah…

Kieran Gilmurray: If all this data and information is available at scale, personalized with intellect and you used the example a moment ago, you know, medical records. Actually, look at the medical industry and think people are so under pressure… It's phenomenal. And if I had a digital assistant as a medical professional beside me, you know, looking up the latest research, finding the latest tablets, scanning
every piece of patient data that's possible in there and augmenting me and helping me make better decisions or even front end, getting lots of questions answered by patients or healthcare or whatever else,
and organizing all those things, you know, it might actually protect me and save me from getting sued quite a few times.

But how do you ensure, Tom, that, you know, organizations use ChatGPT responsibly or other similar large language models inside of their organizations? You know, so the responsibility side of it? And then how do they measure whether they are being responsible, effective or getting an ROI out of this type of technology?

Tom Allen: Yeah. So I think it comes if it’s my personal opinion, and it’s my personal opinion... I think it comes down to, and I think you might agree with it, Kieran… it comes down to the business and the business leaders and making sure that they've got the framework for success in that business. I mean, you can have any tool, it might not be ChatGPT, and that's why I think we're getting excited about it. But ten years from now, God knows where we're going to be! Because this is just, I always say, like everything, leaps and bounds, and this is like a breakthrough that's going to lead to.. in a year... it will just be a standard plug-in as part of your software stack or whatever. I honestly believe that. And people will build on other things.

And another five years, something bigger will come, and I think it will always come down to making sure you've got the right training and the right processes. Would you trust the doctor? And I mean, if a doctor came up to you and hadn't trained for 20 years, would you trust him to give you surgery or would you trust him to do a diagnosis? Probably not! And that's why those fields are regulated, like lawyers and doctors and people, even masseuses; they need to get regulated and get checked. And I think it's the same with businesses. They don't if you don't train your people, if you don't train your staff, if you don't invest in understanding these new technologies and implications, or whether it's data protection or how you handle your data or it's kept, or whatever it might be, you're probably not in a very good spot for using something
like ChatGPT, because God knows, I mean, you see the problems time and time again. I mean, you saw it with Wirecard, a big Netflix thing. You saw it with Enron all those years ago. There are always companies that are massive… And that one company, Elizabeth Holmes, started with the blood-analyzing software.
Time and time again how poor leadership can lead to bad decisions. And they can be big billion-pound brands that get it wrong. Big billion-pound brands get the hiring wrong. As you see at the moment, all the big tech firms are laying off staff. And I think going full circle with what I'm saying about diverging is if you're bringing someone like ChatGPT, the only way is if you're a responsible company. If you clearly look after what implications it has, it's going to be a successful tool for you.

I mean that it comes down to it. If you have people that are just left to their own devices and you won't train them on how to use it, how to look after customers, how to do all these things, how to be creative, how to use it... And you mentioned a very good point, Kieran, about the medical people, with the NHS or private institutions, whatever it might be, it's not letting the software or ChatGPT make the decision, but it’s using it for research. And I completely disagree that ChatGPT should be writing all these articles. And I've seen people boast that it's writing, and I think you are probably going to be flagged someday, or I'd be very cautious that one's going to get under the radar; someone's going to scoop it up from four years ago, and you’re going to be in big trouble. I mean, you're saying your’re of Man City and football at the moment, they're only just coming to… Those kinds of things can lurk there, and then all of a sudden, can set off a spiral chain of events if you don't properly look at it. So that's why I'm personally very against using it
solely for creating a full article and just posting it regardless. But 100% use it for research and then use it for those areas. And again, that's why I'm 100% against it making decisions at the moment for places like insurance and medical work where it can have God knows what repercussions. And we've seen it time and time again where you can have horrible things but make a decision. But it's 100% up for business use. And if people use it in the correct way or try to use it in the right way and use it, I think it's a very powerful tool; I don’t think it’s a bad thing to use.

Kieran Gilmurray: Well, see some of the stuff you mention there, though, Tom… You know, the usual scandals and even my term, the usual scandals, this is if I accept them… I don't accept them at all. Yeah.
People tend not to check or challenge. They get excited by all this tech. So yeah, it poses the question, is there a real threat or an imminent threat to the fabric of society? You know, people are just not managing these tools effectively, not putting the governance in place, not assessing whether they're effective or not,
or even when you know if something goes into the system and it's popped out four years later. Now, technically, is that someone else's copyright or content? Or go back to my example where I said maybe people's jobs would be massively impacted, and as much as we're talking about this, other people aren't talking about it and aren't aware of what's coming. Maybe our education system isn't set up, you know, for a generative AI world or an AI, to get us ready to contribute to an AI world, whatever that means. So, you know, this sort of appears to be societal threats as much as opportunities. What are your thoughts?

Tom Allen: Yeah. I mean, it's got lots of threats. Every piece of new software brings a thousand new challenges and a thousand new potential things, like… the power of iPhones opened up the power of hacking, say, that it's never going to go away. But I think Gary Vaynerchuk did a great interview, and he said he looked at… and they were looking at web free, and they were looking at VR, and it was saying, "Oh, it's opened up all these avenues towards discrimination and racism, and horrible things". And you say, "Well, that's always been there, and it's always going to be there." And he was being very frank with it because he was saying people can pick on the bad things, but it's also… it comes from people. And at the end of the day, you can blame the technology all you want. We had people to solve it and people to create it, and it's only working on past data. It's like if you get ads for certain things in your Instagram feed, it's a combination of what you've been searching for. It might try and recommend you and do data points to spire you off, but people are saying it promotes hate or promotes these things, which ChatGPT potentially can. I mean, I saw an example, and I think CNBC posted it around. They were trying to get around the firewall that Google and Microsoft put up to stop it from giving out any harmful speech or anything. And they got around it, and it showed you how they could eventually filter it to turn into a very horrible machine, a very horrible ChatGPT. And it's a great article, which I think was released about two weeks ago.

So, issues are always going to be there. And I mean, it goes back to our last point, right, making sure that there's always going to be bad people in the world. As much as I’d love for there not to be bad people, people want recognition. Some people's way of getting recognition is for doing very horrible things. And I think it's in business with companies that have mostly collapsed and shown that they're not… they weren’t the leadership they thought they were right for people on a personal level. So I think there are a lot of problems around that.

I think the biggest problem I see is the sensitivity to bias. And that will always be the big thing because one bad fact can spiral very quickly. I mean, look at the elections… Facebook, and I think it's Facebook or someone's being pulled up on it… potentially Twitter or Twitter in the press at the moment for holding back
on disclosing certain information. And they're getting an earful about that. And I mean, it's always going to be the thing. And I think ChatGPT is going to accelerate it. But it's also going to accelerate the great things as well. And I mean, naturally, you can say how bad we are as people, but we live in a pretty awesome world if you look at it… compared to how many people are there and how many people are doing it.
So I think it's, yeah, it's going to be a lot of things, but there's a lot of things to worry about. My big concern is context, truthfulness, and factuality, and I build on those kinds of things. So yeah, it can be great. But I would also just be very concerned about it. And at the end of the day, it’s with the people that are going to use it; it's not, it's never the technology, as you said, the technology that runs on machine learning and reinforcement learning. And maybe we'll get that “I, Robot” situation where it does self-learning. And I mean, you saw on the Facebook bot when they started chatting to each other, and that's a potential reality ten years from now… But I think it's a great tool. Just some people are not going to use it in that way, unfortunately. Probably not.

Kieran Gilmurray: I sometimes wonder whether technology is presenting a mirror to us and we don't like what we see, but we're keen to blame the tech as opposed to ourselves. That said, as a famous newsreader in the UK once said, “You know, why do we constantly focus on the negative news, not on the positive news?” Every time they… in the UK, it's usually 10 p.m. at night when it comes on... So again, interesting times.

So if we finish on a positive question, Tom, maybe, what do you see as the future of conversational AI? What would it look like in, you know, next year, five years, ten years?

Tom Allen: For me, five personal chatbots that are doing everything for me that I don't like doing. But I don't mean… I'm joking about that. I mean, I would love a person. I think it would be very powerful to have it. And especially for things in your own personal life, booking things and getting things done, and reminders and scheduling assistants can be very powerful. But for five, ten years, I mean, I always say… people think I'm crazy because I think space travel will be possible… because I just think it's like when the horse turned into the car. Henry Ford is a classic line. “If I'd ask people what they wanted, they’d say faster horses”… it’s, you know, you have those people, and I think ChatGPT is that strive towards a new era.

I think Elon Musk, whether you love or hate him for what he's done at Twitter, what he's done with Space X is just incredible because he's commercialized and opened it up past what used to be just governmental contracts. And I can see ChatGPT opening up all sorts of new industries. I mean, you look at, I don't know, this kind of popped out of my head, but I saw something: the amount of people starting their own businesses today compared to 30 years ago is ridiculous. Like so many people have it, and so many people are creating side gigs, and so many people are creating the creator economy that I think ChatGPT will accelerate that, and it'll bring a lot of positive change, a lot of positive things. And if occasionally people have to be brought up for using it the wrong way, so be it. Because, you know, every challenge is a lesson, so people can learn from it. And sometimes, it's at the expense of a company or a person. But I'm not saying that's a bad thing, but it's ultimately going to be very powerful in ten years. And I always go back to what I think. I think within my lifetime, we will be going to other planets because I think all it will take is one small manoeuvre and ChatGPT.

That's why I say it's hype at the moment, and rightly so. Hype like the internet bubble was a hype. I mean, certain people said it wouldn't last and it won't be around. And there's like there's an article in The New York Times saying like in 1998 Internet is a bubble, it's a hype, it's not going to exist.

So it just shows you how wrong it is.  And I think ChatGPT is the door to a new way of potentially 
taking over websites or potentially looking at digital infrastructure. I mean, now, when it used to be that you'd cold call people, now you WhatsApp businesspeople like… it's just another change in how you do business and how you do things. And I think we'll have a lot better singers, we'll have a lot better film directors, will have a lot better people in business operations, a lot more people that are younger potentially coming up with more creative strategies at a quicker rate. So I think it's a big, big positive. But as we said, just be careful of those biases. And the main thing I have is, is if being factual and being an honest fact checker. So yeah, very positive and very, very I think we're just getting started… where we are, just getting started. God knows where we will be in 10-15 years.

Kieran Gilmurray: I know it's kind of exciting and kind of scary, isn't it? But I do love the idea of a digital assistant just allowing me to be more productive in less time. And I think that's the balance for me. You know, I don't want to be doing 40, 50, 60 hours a day with a digital assistant helping me double and treble up. But I wonder if we can get some time back or maybe, as we mentioned earlier on in the medical profession, you know, actually, resource augments medical professionals with this type of technology 
and not replace it. To allow us to make better, clever decisions. But as you said earlier on, old tech creates more jobs than it destroys. The key is getting ready, though.

Tom Allen: Yeah, and if you’ve seen the film or if anyone watching this has seen the film “Her”, which is what, 2013 with Joaquin Phenix? It's a brilliant film, a little bit explicit at times, but it's a very good example of when I watched it, a very good example of what we've got to come, and it was spot on with what we see with ChatGPT. So you can look at all the bad things, but look at mental health. If you have something… or look at the follow on from booking appointments at the doctors,  or looking at the follow on
from the insurance or having that kind of… it can make a brand so much more personable, which I think is a big thing that people are really racing against. Because if I can get those updates on annoying me but kind of nudging me in a more personal, friendly manner that I can kind of put a face to a brand... it’s like going to the airport, and you see a hologram I saw one at the Birmingham Airport recently, and it was like, “Get your passport ready!” in a very friendly, approachable person. And that eventually will switch that you'll just have on your phone.

So yeah, as you're saying, it's going to be phenomenal, the world we live in, and hopefully, I get to see it.

Kieran Gilmurray: I hope so, too, for you, and I get to enjoy it. Tom, founder of the AI Journal and AI enthusiast. Thank you so much indeed for coming on to DRUID Talks webcast today. Folks, that's all for today!

As you heard from Tom, this is a very exciting technology, a very exciting space. It's not without its imperfections. And that's true about any technology in the world, as we mentioned earlier on. But certainly, the opportunities that this and conversational AI can create for businesses now, next year, and in ten years’ time open up a whole world of opportunities. As Tom says, let's hope we're all around to enjoy those opportunities as well! Until our next webcast, thank you very much indeed!