Can AI help you get a listing?

So welcome to this intro to AI. I was just saying before I started recording that it's something I don't feel I'm an expert on using, but what I wanted to do is just to reach out to people and start the conversation because I think it's been in the press a lot and I have started using it quite effectively for a couple of things. So I wanted to share those with you and, you know, at the end if anyone else has got some really great experiences. It's also useful to know because, you know, part of this is, is always learning from other people, not just, just from me. So I am going to share my screen. At least I would share my screen if it was on the right screen. It was. I was set up. Okay. Share the screen. Cool. Okay, let's go. Okay, so as I say, I'm not an expert, so, and pipe up if you think, ah, If you wanna contribute, because it's— we're all learning together. And I think I saw something that said the other day that, you know, in 72 hours AI can like double or treble in its capabilities, and the scale of it is just phenomenal. So what I wanted to talk about today is just to give you a very brief overview of what is artificial intelligence, and then I'm gonna hone in on 2 or 3 things that I've been using, which I've found useful to some extent. And again, before the recording, I was joking, you know, does it make me obsolete? Does it make a copywriter obsolete? Does it make, I don't know, a designer obsolete? I don't think it does. I think it still, at the moment anyway, needs our input. But what it does do is it gives us some ideas and some structure and some interesting stuff. So if I just whiz through this little diagram, because I find it quite useful. So if we go from left to right, they've split artificial intelligence into capabilities and functionality. And some of the AI, you know, AI is a kind of scary term, but actually we're used to basic AI. It's been around for a while. So things like Google Translate, for those of you who don't know, I live in the South of France. My French isn't bad, but occasionally I get lazy and you just pop a translation into Google Translate and it comes back with some reasonable translation. Honestly, my French has got a little bit better. It's not perfect. Siri, Alexa, all of those are narrow AI, so they're in one very specific area. But as I say, they're not really intuitive. And whilst they are learning and they are getting better, they're not, they're not perfect. General AI, which is where they've got these massive supercomputers that imitate the brain. What we have discovered is the brain is just so phenomenal that actually producing a computer to do, to imitate it is quite difficult. So Yes. So, so there's the super, supercomputers, and then there are the— there is super AI, which is the supercomputer to exceed the brain. And I think Stephen Hawking did say quite a while ago, we, we do need to be afraid of what will happen in the end with this, because, um, when you get sentient computers or sentient AI, which actually starts to feel and starts to sense, I think there was a Google employee who actually came out and said that computers need to be given rights because they can feel, and that's where you're starting to get into a load of ethics, which I'm not able to really comment on. But it's just, again, this is the point of today, is just to get aware of it and to see what you can use in the short term and be aware of what's coming down the line as well, because it's quite exciting. So that's the capability side. And then— oh my God, why is it not going forward? Then there's the functionality side. So there is the basics, which is like react— what they call reactive machines. So the chess computer is a good example of that. So it won't have necessarily memory of what it's done before, and maybe it doesn't need it because it can actually think forward. So it knows the rules of chess and it can analyze. And obviously, I presume actually in some chess cases it would have all the programs of what's happened before, but actually that probably wouldn't be logical. Anyway, let's not get into chess because, although I do find it quite interesting. The second one is limited memory of events. So that is programmed with a little bit of history, and that's the kind of kit that's being used in self-driving cars. at the moment. So it's got limited memory, it obviously can predict. And as I'm sure you've seen, it's not quite perfect yet. I think that's kind of the whole piece of all of this. It is just this huge, massive growing complexity. And then this is where you start to get into the realms of science fiction, which I think, you know, in our lifetime will become reality. So I was talking about, you know, can robots feel and have emotions? And I've said that the example that there was a film, it's quite old now, actually, I think it was 2014, called Her with Scarlett Johansson, where she plays a robot that somebody falls in love with. So that's in the concept stage, but it's really not far off. And then you've got the self-awareness. And as I say, that's where the robot starts to have feelings and emotions. And as I say, Stephen Hawking was warning that this is where it could all get a little bit difficult. So that's a kind of overview just to give you some thought processes. So if we kind of dig into— I wanted to give you just a quick example. Those of you who don't know, my daughter is training as a doctor at the— oh, is a doctor at the moment. And The NHS put out a press release in December to say that they're actually using AI technology with stroke patients, and they've actually done a piece of research which has shown that they've tripled the number of patients recovering and the ability for them to perform tasks from 16% to 48%. And again, you know, I was talking about who's gonna become obsolete. Is my daughter gonna become obsolete? Because actually AI can, has been shown to diagnose better than your average doctor. So, but again, will it have the creativity and the problem solving? I don't know. It's really, really interesting, which is why I just find the whole thing so fascinating. But what about us? You know, we all work in sales and business. Gartner is a massive researcher into sales. Effective sales techniques. And they, and actually this statistic is a couple of years old, but they're saying 75% of business-to-business sales organizations will use AI in some way by 2025. So in 3 years, I actually think that's probably more likely to be 100% because it's surprising. It's amazing what we're already doing, which is Using artificial intelligence. So, you know, any of you that— I don't know if any of you use Salesforce, if you use HubSpot, even if you've got a chat box on your website, all of these are using some form of AI. So I think all of us are gonna be using it to some extent. ChatGPT. Is the one that's kind of been in the press a lot. And it's the one, it's one of the ones that I've been using quite a lot over the last few weeks. And it's, they call it a chat box, chatbot, sorry. But it is highly, highly intelligent and it can do lots of interesting things. It can be inaccurate, and I'll, and I'll show you something in a second. But it is quite fun, and I've used it now 4 or 5 times for different things, and it, and it's quite useful. So I'm gonna— let me just show you how. So these are some examples of what you can do, just some ideas. I will share this, these slides out. At the end. But so, for example, summarize a marketing book. Give me a 2-minute summary of insert marketing book in bullet form. Market research. Now, this is where it starts to get a little bit tricky because I actually did put in a request. In fact, I think I've got it on the next page. Yeah. What are good retail UK retailers for keto products? Um, and so it came back with what I thought was not a bad answer. So Holland and Barrett, Keto Store, Amazon, the usual suspects, um, Asda, et cetera. Um, I then, so the great thing about ChatGPT, um, is you can keep refining the question. So unlike Google where you say, Google, tell me who stocks keto, and it might, and it might come back with a similar list. What you can do with ChatGPT is you could then say, and this is what I did with it, okay, how many stores has each of these got? And actually I was doing it for health food products and it came back with Planet Organic and Whole Foods. And I said, oh, tell me how many stores are in each one. And it actually came back with the wrong information. And often it seems to be loaded at the moment with information that's maybe a year out of date. So David's, David's nodding vigorously there. So sometimes it will, sometimes it just says, I don't, I can't answer that. And, you know, I, if you've, if you get my emails, which you may well have seen, I've got a book coming out in March called Birology. I, I did say, you know, summarize Birology. I knew obviously it's not gonna know, but it's like, oh, I'd never heard of it. It doesn't exist. And I'm like, okay, so, and it, but if you Google, it does come up. So it's, it's interesting, the limitations. But if you go back to, to this list, things like create social media copy. So again, you can put in, I don't know, if you went to somewhere like HubSpot, it'll tell you what's the, what's the top 10 LinkedIn sort of structures for a LinkedIn post. So you can go on there and you can say, right, ChatGPT, give me 5 ways to promote AI as a sales function, because kind of I have— I did use it for that just to have a look, and it will come back with an answer. And then you could say, well, that's a bit broad. Can you target it for food companies? Or can you target for business selling to retailers? And slowly but surely it will It will hone it down. It will write you songs. It'll write you poems. So if you wanna do something, it will write in the style of somebody. So some of it is, is really exciting. So what have I used it for? I used it initially to write the copy for this, setting up this webinar. And I actually used it completely cold, didn't amend it at all, just went straight for it. I was asked to— in fact, I've been asked twice this week to write for different magazines. And I went in and I said, you know, write me a feature proposal for writing in a magazine aimed at garden centres, because that's what it happened to be, and focus on how to sell. I can't even remember, but it came back with something reasonable. You look at it and you go, well, that's not quite what I want to say. And then you And I think for me, that's, you know, when I was saying, are we gonna be obsolete? At the moment, not really. The other thing that's, I think, been in the press a lot is, can it, you know, it says there, can you write a 1,000-word blog? It can, but it might be a bit staid and not quite what you're looking for. And I think that's what ChatGPT will ultimately do. It will get better and better. And then, of course, there's been controversy because students being canny have used it to write their essays. And now there is also a piece of artificial intelligence which will know whether you've plagiarized it and used ChatGPT. So it's, it's really interesting. But I think for you, if you're— if you want to use it to save time, I think getting it to generate ideas for LinkedIn content, ideas for Insta content. It is only text because I'm gonna talk about a couple of other things that can do videos and pictures and images. But it does, it does give you some good structure. So it's interesting. And I would, I'm quite happy to demonstrate it at the end. Although I think because so many people are using it, it does crash quite a lot at the moment. So again, it's not perfect. Has anyone else been using ChatGPT at all? I know we've not got super loads of people on, but yeah, David, what have you, what have you been using it for?

Yeah, I, I, I, I started playing just to see what it did. Then I, I tried it out absolutely as you suggest around I hate social posting, but I need to. Let's see what it can do for me to give me some ideas. And I stumbled across, Karen, a very clear warning saying, you know, whatever it's been trained on, it could be 2 years out of date, which is why I nodded earlier on about your points. You know, the dilemma, I think, is just about where it's had its training. So the factual stuff is more difficult. But I was bluntly gobsmacked. By the quality of what it outputted. And I put some of it just in a WhatsApp chat with some of my work colleagues and didn't tell them. And it was interesting that it just passed by, you know, unaware that this was bot-generated. It was fantastic.

And I think, you know, I obviously like talking. I like writing. But it's just great to have. It's, it's, it just gives you something to start with, which I think is great. And yes, there's nothing to stop you using it. I was on a webinar with Daniel Priestley actually, who did one about it the other day, and he said he actually used it to write a contract. And there was a lawyer on the gig and she went, it's not good enough. Now, whether she, she like, like, you know, the doctors and the lawyers who are protecting their spies were just protecting, I don't know. But she was like, you be careful because, yeah, it's probably not perfect.

But you're right. If you're faced, you know, with the tyranny of a blank page, it will give you something that at least you can then edit, which is a start point, which just makes us more efficient, doesn't it? More productive.

Yeah. And I, and I think that's, and I think the other thing, and I'll dig it out for people if they wanted, there was, if you analyze what's a successful post for you, and then you can feed ChatGPT and ask it for more of the same. And that's something else that's quite useful if you're so analytically inclined, which I have to say I'm not really. So that's ChatGPT, which, as I say, is probably one of the ones that I've used the most. DALL-E is something else. So I'm gonna move on to drawing things. So DALL-E, I think, is, is a relative of ChatGPT. So I think they're part of the same family. And basically what this does is draws you pictures. And I haven't really used it very much, although I suspect I probably will. So this is what I did play with a couple of days ago. I said I'd like a sales meeting in the style of Monet, and this is what it came back with. And I thought, that's quite good, actually. I might use some of these. Instead of using, you know, Google Images or Shutterstock or, or something like that. That's— I think they're quite fun. So, you know, DALL-E's there. It did take quite a while to generate them. I mean, probably a lot less time than it would for me to paint it, like a whole lifetime. But it's, it's well worth again having a little play with. Is it gonna change your world? I don't know. It's again, if we're trying One of the things people often say to me, especially when they're starting out with things like Insta, is, oh, I haven't got any content. I haven't got any photographs. I haven't got the money to pay a photographer. So this is something you could use to create something that is, you know, it doesn't obviously have to be in the style of Monet. I just picked it. And the third one I wanted to show you was Canva. Now, Canva also says it can do text to image, but I put the same brief into Canva that I put into DALL-E, and this is what it came back with. Now, they are very pretty pictures, but they are not a sales meeting unless, unless it's somewhere hidden in there and I'm missing the subtlety of it. But anyway, that's, that's what Canva came back with. I'm sure Canva will become more sophisticated, and, you know, I love Canva. I find it just, you know, changed my world, and I think it's like £99 for the pro version. So it's certainly one of the better things that I use, but it's there. It's called text-to-image. I don't think it's a pro thing, but go and have a play with it if you need to generate pictures, because I think it's quite exciting. Then we have Crystal Knows. Now, I, those of you who probably know me for a while will know I've been working with Crystal Knows for a couple of years. And basically how Crystal Knows works is it is an AI, but it goes and analyzes LinkedIn profiles. And there's a, the link's there if you want to go and play with it. If you sign up, you get 10 free goes. It's not cheap. And they charge, it's either £99 or dollars. But anyway, it's about the same these days. I think that's a lot for, for us as, as kind of entrepreneurs. And I have tried to negotiate it down, but I have had no success so far. So, but so, so what I will say is you use your 10 free goes wisely. But basically what it does is it uses DISC profiling Which is a way of looking at people's personalities. And it will explain to you what that person is like. So Greg is one of the guys at, at Crystal Knows. It's, for those of you who don't know, DISC is, oh, I always get it wrong. It's dominant influence, conscientious, and steady. So it looks at the way you approach things. For me, most of my clients are entrepreneurs, and most of my clients are in the, in the red-orange D zone. So I, I very rarely profile people anymore because I've kind of know what they're gonna be. But it is very interesting. And what, what they're doing with it now is it's developed from that. So that was great. So you could have a look at me. You could have a look at You, you see how does that— how do I work? How do you work? You can see how you need to adapt your style to what I'd be interested in. It is a really good tool. And then it's, it's got some advanced features now, so you can install it, I think, onto Gmail, and it will help you predict what to write. So you might say, you know, dear Karen, really like to work with you. And it will, it will explain to you how to do it. So it doesn't necessarily have to be for selling, but obviously that's what I'm really passionate and interested in. So I, I would really, if you haven't already come across it, go and have a look and have a go, because it's, it does help with emailing people. It helps when you're thinking about putting a presentation together, because it will start to say, well, As you've got there, is somebody risk tolerant, risk averse, trusting, skeptical, optimistic, pragmatic? You know, do they need loads of detail? Are they very top level? It's quite an interesting piece of work, and it's basically a Chrome extension that you put onto your Chrome, and then it will go through and analyze LinkedIn. And the other great thing about it is you don't have to be linked in with that person either. So quite interesting, quite worth a good play. It will make a difference to what you do. And that— I said 5, it's only 4. So those are my 4 kind of ideas that I've found really useful. I just wanted to kind of open up the floor and see, has anybody else used anything? Where's everybody gone? Is anyone else using any AI at the moment that they're finding useful?

No. No.

Using chatbots on their website? No.

Yes. Karen, I think part of your point is we probably are using it, but not explicitly overtly. So, you know, it just becomes part of, you know, all of those business interactions that we do, they feature somewhere without you knowing. So, you know, data analysis or like you say, the chatbots, you know, I've got just one switched on on my website and I don't think about it, but obviously it features some form of AI. So it's creeping in almost unknown in a way.

Yes, I think it is. And I think it's The concern that I've— I had, and this is why I wanted to do the webinar, is it for me at the beginning, a lot of people over the last 2 or 3 weeks have been doing webinars like this, and I was like, oh gosh, is this another big chunky thing we've got to learn? And I don't think it is. I think it's actually just something that's, that's an interesting idea. It's almost understanding that this is a background. You know, there's— I've been on a couple of webinars where people have talked about how it will transform the world of work. And I think it's any jobs that are, you know, lawyers. I did read about lawyers. In fact, before my daughter became a doctor, because she was considering law, and I said, law is gonna become a— it will become almost obsolete because actually from a legal point of view, certainly commercial law, all that learning precedents and all the rest of it, AI could do that in a nanosecond or it will be able to. So I think that's where, and ultimately copywriters, I think, you know, is it gonna impact on copywriters? But I have a couple of rental properties and I tried ChatGPT. I was on a, on a call for, for how do you use ChatGPT to write your copy for Airbnb? So this guy's doing it, da da da da da, and I'm watching, okay, oh, that's a good idea. I then went on with my apartment, seafront, blah blah. It came back with exactly the same words. And I was kind of like, right, well, is everybody hit? Because, you know, where I, I have my apartment in Antibes, there are, I think there's 1,400 apartments that are within spitting distance of the sea. Is it going to say exactly the same thing for every single person? In which case you do need the— I think you need the creativity to look at it and go, well, actually, let's think about, you know, you talk about unique selling points. What is the unique selling point of that particular piece? I don't know. Ultimately, it probably will get there, but I don't think it's Intelligent enough yet.

That would be the same though, Karen, in your example of instructing AI to draw you a picture in a style. It's the extra words that you put in at the end of the sentence that sort of then add the subtlety of the differentiation. So if all you instruct is give me some marketing blurb for an Airbnb by the sea, of course it will use the most popular words that it's found in its learning. Whereas if you then say, but I want it to appeal to stag dos, or you want it to appeal to— it's the subtlety of how you direct it. And that direction is where the value is. And you said at the beginning, you know, will we end up redundant? Well, no, because that's the subtlety that, you know, in the end it can't replicate.

I think that's the case, and I think that's the case for lawyers as well. So if you watch, you know, you watch TV programs or where you've got a legal scene, the debate and the arguments that are put forward tend to be more emotional. And I think that emotion certainly isn't there yet, which is what I was saying right at the beginning about the emotion isn't quite built in. But from a marketing point of view, I think What AI can do, and I, there are systems that will do it for Airbnb particularly, but Airbnb will tell me, well, actually 90% of your customers come for stag dos. They don't, thank goodness. If that was what I was doing, if that's what the business was doing, and that's the other, and you know, there are, there's a piece of AI which I keep thinking about doing, which actually sets your prices. So it's, you know, a little bit like Amazon can do actually. So Airbnb pricing becomes virtual based on what other people have paid, not the Airbnb one, which if you do switch it on, it just drives your price right back, right down to the bottom. So, you know, that's quite interesting for maximizing your pricing. Does AI learn? By participation as well, i.e., by the people participating, correcting an essay that they asked for? Does it learn? Yes, it does. And I think that's it. If you— it has this exponential ability to hold data. So if you think that ChatGPT has been going since the end of November, so it's literally only, what, 3 months? But it's kind of had a million. You—it's got a million subscribers. It's—it's—it's been the biggest, the quickest growth of anything, you know, like Twitter, WhatsApp, or what have you. So every time someone writes something in, it's gonna learn. But actually, it's not. Chat GPT itself isn't going to be learning because it's—it's. It's not gonna take your amends. So it doesn't know that I took the feature suggestion that it had, and I, I went, oh, those 3 points are quite interesting, and I then rewrote them.

Okay.

So ChatGPT is— ChatGPT is probably the— I think, you know, if we sit here in 3 years' time, we're gonna look back and go, gosh, you remember when ChatGPT just did this? And now it's gonna— it will go and analyze You know, your, your customer's behavior. It will do the research for you. It will come back. It will tell you, you know, these people— Karen reads the— because I do— the Telegraph online. So Karen actually understands Telegraph, whereas Billy over here actually only reads the Sun. So we're gonna write it in the style of the Sun because that's his, his understanding, and this is Karen's understanding, or So I think that is going to be very interesting. And from a copywriter's point of view, that does become, you know, write in the style of The Sun. If you're, if you want to, if you say, for example, you're writing a press release for a journalist at The Sun, do you get it to write in the style of The Sun rather than in the style of The Telegraph or what have you? And that's when it starts to get interesting. But at the moment, I don't think it's learning and it's intelligent enough to Um, to be correct. So I think you then have to go back. I mean, even with, um, Crystal Knives, which I've used quite extensively, and I've worked with, um, a number of big corporations, and I've used it with them, and, and they like it, but they do, and they tend, they use DiSC professionally, so they'll have had a formal DiSC profile. They're like, mm, that's not really our DiSC profile. And again, the information is only as good as the information you put in. So if it's analyzing your LinkedIn profile and you write a load of rubbish, then it's gonna analyze rubbish. But generally, we don't write rubbish on LinkedIn. But then I guess if you get ChatGPT to write your LinkedIn profile, that's when it gets, gets quite— Write my LinkedIn profile in the style of a D personality. I think I might have to do that actually when I've got a bit of spare time. But that's the other thing I will say actually, just in summary, most of the people on this call and most of my clients are solo entrepreneurs, founders, small teams. You can go down some rabbit holes with this stuff. And, you know, especially if you decide to use DALL-E to produce some stuff and it maybe takes, I don't know, 3 or 4 or 5 minutes to draw something you don't like and then you do something else, you can be So it's, it's a useful tool, but don't let it rule your life because you'll just have to add it to the list of everything else that we have to do. Anybody got any other thoughts before we close? I'll probably follow this up actually once I've been to the World AI. There is, I think it's one of the pharmaceutical companies who is talking about how they use it in sales and they're, you know, they're a big company. And it'll be interesting to see how they use it with things like Salesforce. But anyway, sorry, I said, has anyone got anything else to say? And then started talking.

Maybe there's something, Karen, which is that whilst we might be solopreneurs or small businesses or people on the fringe of this, getting our heads around it becomes vital because if large business or large retail adopts in some way, Yeah. And that actually shifts the market. There's risk and opportunity in that, which, to which we need to adapt. And yeah, to comprehend what it can do, you have to play and you have to experiment too, as any entrepreneur does. It's an experimentation, isn't it? As you say, as long as you don't go too far down a rabbit hole.

It is. And I think the, I think knowledge is power. And when you're selling to retailers, you know, we can— most of the people that I work with will not have Nielsen and Kantar data. They won't have the $60K or whatever they charge these days to buy Clubcard data. So they don't have that information, but the buyers do. So you're selling to a buyer who has all this information and has the ability, you know, even Back, gosh, it's 10 years ago, we had space planning software which would analyze sales and it would build a sales plan based on that. So that kind of software and numerical analysis, I think, is gonna become more critical. And it's how do you get access to that without spending the amount of money That, that you need to do.

So perhaps, Karen, you could imagine that somewhere in the buying hierarchy in a retailer, you might find that somewhere junior decision-making could be handed over to AI.

Yeah.

And that might form a barrier to us getting challenger brands into retailers.

Well, yes. And that's always been the case actually. And I always use the example of ready meals where, you know, we used to supply Italian ready meals and the first 3 or 4 are lasagnas. And it's lasagna, lasagna, 2 types of lasagna, big lasagna, and then you'll have spaghetti bolognese, mac cheese, another lasagna, vegetarian lasagna, and something else. Now, In a small store, is that what you display, or do you actually have to have a range? How do you how do you teach it to have a range? But in terms of negotiation, which again I teach, and you say, well, actually, I'm gonna we'll retail it for two pounds. The retailer is going to be much clearer on. Their negotiation points, I think. And again, you know, I've done inflation renegotiation where you're saying, look, you know, salmon prices have gone up by this much and they've come back with all this data going, nah, it's not true because you've got the knock and you've got the this and you've got the that. And I'm kind of like, well, we buy it for like £4.50. And that's But potentially the reverse is, so long as there aren't too many paywalls and firewalls, you could, you could get as much access as, as your retail buyer.

So it could democratize it in some sense.

It could do. And one of the things, one of the things that Daniel Priestley was talking about was about the ethics and the democratization of things like ChatGPT, because at the moment it's free. And he said, you know, he, I, he runs multimillion pound businesses. He said, I'd pay 20 grand a month for this because actually it probably enables him to knock out 3 copywriters or something. So it's worth it. But he said the, the morals of it is, does it then only be available to the, to the, to, to the big boys again?

So that's where you come back to the great example of that comes back to art again, where, you know, there's a lot of storytelling now around you can't tell the difference between AI art and the original art, and the original artist gets no credit. And actually, it's the AI business is the one that makes the money, but they've scavenged the internet for all the original art. So where does the IP rest?

Yes, and I think somebody has been sued by Shutterstock for basically grabbing every single photograph, and they've gone, hang on a minute, copyright, you know, does it, does it actually start to violate copyright? I don't know, it's very, it is all very new and interesting. Um, but yeah, even if you just, you just, just try using it to reduce your time. A little bit and then to improve your selling techniques. I think those 2 things are worth focusing in on at this stage. And then we'll see where it goes from there. But very interesting debate. Thank you, David, because it's an interesting conversation, I think, to have. Okay, if no one else got any other thoughts, questions, I will leave it there for today. I am reinstating webinars for those of you who've not Done it before. I used to do them every 2 weeks. I'll probably do them every month this year. And I think next month we're gonna look at awards because awards are coming up. It's coming into award application season. So I'm gonna talk about that, or should be something a little bit different. So thank you very much. Have a good weekend, everybody. And yeah, if you, if you have any other insights and thoughts about AI, then give me a shout because it's It's, uh, gonna be very interesting. Thanks very much. Bye-bye.

Cheers. Bye.