So if you say to somebody, here’s a machine, you can have a conversation with it and it will guide you to the solution to your problem, it will help you, right. It will help you code. It will help you. I don’t know what else it does. It will help you write your emails.
It will summarize your emails. It will make your day-to-day life easier. That’s an additional service that we didn’t have before. Search doesn’t disappear because this new service has appeared. What does happen is that people still use search, but they use ChatGPT Microsoft Copilot perplexity.
John Corcoran: 11:38
Interesting, interesting. And I don’t know what the datas are from this year, but I know that I saw that Google’s traffic in 2024 last year went up by 20% over the previous year.
Jason Barnard: 11:53
Yeah. And SEMrush just did a huge study where they’re showing that search has gone up alongside use of AI-assisted engines, as I call them. I think what is true today is people are new, people are being brought into the arena, and the people who are already in the arena are suddenly think, well, I can do this and this, oh, and this and this and this and this and this. And they’re expanding their usage. What will happen is people will gradually get used to the idea that search is pretty clunky.
I have to actively ask a question, go through the links, figure out which one I want to look at, and little by little, people will start to realize you can just ask ChatGPT or Google AI mode to summarize.
John Corcoran: 12:44
And so at.
Jason Barnard: 12:45
This point.
John Corcoran: 12:46
And that point is one that has been fiercely debated. You know, there’s a story recently, Cloudflare came out and said that we need to change the economics of the web because publishers are not going to get traffic. And for the longest time, Google’s promise has been that we’ll send traffic to these websites and then websites and publishers and service providers are, you know, fearful that they’re not going to get people that are coming to their website because they’re inquiring about a particular service. What are your thoughts on all of that? How’s that going to affect publishers, and how is it going to affect the economics for people that have websites that have been putting out content for a long time?
Jason Barnard: 13:29
Well, if you’re a pure publisher who is publishing content like USA today have famously now just blocked. That’s your business model. You’re in trouble. However you look at it, however you manage it, you’re in trouble. If you’re trying to sell something, it doesn’t matter.
Because the point isn’t do I get people coming to my website? The point is, do I make a sale at the end of the day? So you’ve got kind of two schools of thought there or school two?
John Corcoran: 14:00
Yeah, two very different approaches, I guess. Yeah.
Jason Barnard: 14:03
And I think the publishers are getting a lot of press coverage because they are the publishers.
John Corcoran: 14:10
Yeah, sure. And they write about these topics. Yeah.
Jason Barnard: 14:13
So yes, there’s a huge problem for them. But no, there isn’t a huge problem for anyone trying to sell something.
John Corcoran: 14:19
So if you are a company trying to sell something through your website, then do you continue to create content online in the way that you have before, or do you change your approach?
Jason Barnard: 14:29
You change your approach, but the approach that you need to take isn’t so very different to what you had before. The old SEO strategy was, I look at what’s ranking today for the keywords that I’m interested in, and I will create a piece of content that is better than the one that’s ranking, and I will therefore rank above it.
John Corcoran: 14:50
Yeah.
Jason Barnard: 14:51
Made sense, but once AI has the information, it doesn’t need it anymore, so there’s no point in doing that. Because you don’t beat anybody because it’s got the information. It isn’t going to come to your website to look for that information, and it isn’t going to prioritize you over anybody else because nobody gets attribution for that information. So that particular strategy is now dead.
John Corcoran: 15:19
So the strategy of trying to target a keyword and create content that’s better than what’s out there is dead. So what do people do instead?
Jason Barnard: 15:30
You need to expand human knowledge.
John Corcoran: 15:33
What does that mean?
Jason Barnard: 15:35
It means that if the machines don’t know something, or you can create clarification or detail about something that they need, they will always come to you. But if you produce something that is the same thing as everybody else, but slightly better in the way it’s presented, they don’t care.
John Corcoran: 15:54
So like simple things. So like if someone’s searching for it, whether it’s in Google or ChatGPT or whatever, if they’re searching for how do I screw in this light bulb? They’re not gonna they’re no longer going to go to some niche blog to find that information or to a handyman company’s website, which is written a helpful tutorial on how to screw in a light bulb, because that information is going to be found very easily, either in the AI summary in Google or it’s going to be found in ChatGPT. But if you can write about something new and innovative that isn’t there, then that’s where you should spend your time.
Jason Barnard: 16:34
Yeah, and always remember that when you create content that’s top or middle of funnel, you’re probably not going to get the click anyway. So you’re only creating it to educate the machines. And I’ll give you an example: I’m a double bass player. And I have a huge bass amp that’s really cool. And it it makes the best bass sound in the world.
And I wanted to play guitar and I thought, well, I don’t really want to play. I don’t sorry, I don’t want to buy a new guitar amp. So I asked ChatGPT if I play my guitar through the bass amp, will I break it? ChatGPT said, absolutely not, but it will sound really bad. Okay, can I solve that as a problem?
Yes. If you get some effects pedals, you can solve that problem. Which effects pedals do I need? You need a compressor, an equalizer and a reverb. Brilliant.
Which ones? Gives me a list. How much will that cost me? $250. Can you get that for cheaper?
Yes. Are there lower grade or lower quality pedals that, if you’re not a professional musician, will be just as good? And the total cost is $150? Brilliant. Where do I buy them?
If you’re in the US, buy them at Sweetwater. If you’re in Europe, buy them from Toman. I’m in Europe. Give me the link to Toman. Link went through.
Bought $150. Sale to Toman in literally 20 minutes. From the moment I said, can I play my guitar through my bass amp without breaking it?
John Corcoran: 18:05
Right. So then. So coming back to the idea of writing about new ideas, new concepts, stuff that hasn’t been written out before. So how does that relate to that story you just told then?
Jason Barnard: 18:18
Right. So the huge problem is that in the old world of traditional SEO. I would have researched all of those topics. Can I play the bass through an amp?
John Corcoran: 18:29
And you would have probably landed on all these different web pages. Yeah.
Jason Barnard: 18:32
And somebody would have had a YouTube video about it. I would have watched the ad. I would have given them money through the ad. The fact that I watched it, that simply doesn’t happen anymore.
John Corcoran: 18:43
So then is that something where a company should try and write an article about that? Because you the way you described it, all of that came through the LM. There’s no need for. Okay.
Jason Barnard: 18:56
So if your career is creating, in this particular case, videos of you playing the guitar through a bass amp, your career is pretty much over.
John Corcoran: 19:07
And I’m if you’re explaining to someone like kind of stuff that can be explained in an LM.
Jason Barnard: 19:14
Yeah. If you’re creating publishing content, I mean, anything that’s just news or helpful or DIY or and you have nothing to sell at the end of the day. There is a huge, huge problem and I empathize. But that’s the way the world is and we’re not going to change it.
John Corcoran: 19:34
But if you have some kind of service that’s going to help with help, help the the customer to get an end result faster than you may be. Okay.
Jason Barnard: 19:44
Yeah. So, I mean, taking the example of Sweetwater or Tollman, if I create content as a marketer at Tollman or at Sweetwater that answers that question, then what happens is, as the user has the conversation with the machine, my company is top of algorithmic mind through the funnel.
John Corcoran: 20:05
Okay.
Jason Barnard: 20:06
And then when it comes to the perfect click, which is the moment when I clicked on the link and went through and bought which fabrics canal from Microsoft, Bing calls the perfect click. That’s what they’re doing is bringing the person down to the perfect click, I click, I buy.
John Corcoran: 20:22
The perfect click and the idea in the sense that they’re preconditioned to buy. At this point, like your website no longer needs to serve as kind of a conversion funnel because the LM has been a conversion funnel, like you’ve gotten to the point where, like, boom, I’m ready to buy, and you click right through and buy it right away.
Jason Barnard: 20:39
Yeah, yeah. And Fabrice was telling me a couple of years ago, that’s what copilot does. We’ve built Microsoft Copilot to recreate your funnel in our engine.
John Corcoran: 20:51
Yeah, maybe even more efficiently than your website.
Jason Barnard: 20:54
100% more efficiently. I mean.
John Corcoran: 20:56
It could be a really interesting solution because, you know, right now, the smaller providers or sellers have to compete with, like, the really big ones. So if like Amazon has a really, well, converting funnel to sell a widget and you don’t have a great, well, converting funnel to sell a widget, you’re going to lose to Amazon. And that’s what’s been happening over recent years, so that it’s almost like it could hopefully level the playing field for smaller sellers, smaller providers, smaller businesses.
Jason Barnard: 21:26
100%. Well, the whole thing is you need to recreate your funnel in their brain. And if you can do that and your niche and you have an advantage over Amazon, you need to describe that advantage and convince the machine that that advantage is worth it. Recommending you instead of Amazon.
John Corcoran: 21:44
So you do. So you need to create content that is that is around your expertise around your your focus area. Yeah. Okay.
Jason Barnard: 21:53
And and and the the problem that you’re always going to have is that your boss is going to say, well, nobody’s visiting the website. Nobody’s visiting this page. Well, the, the, the role of the page isn’t to get people to visit, it’s to educate the algorithms. When I say educate the algorithms, I’ve been talking about that for ten years. Educate them like children and they will bring your audience to you, but they will only bring the audience today for that perfect click.
That’s what you’re aiming for.
John Corcoran: 22:23
Let’s talk also about gaining credibility through external sources. Let’s call it awards, associations, accomplishments, stuff like that and how you communicate that online. So for companies that are out there that want that credibility or want to communicate it for the pieces that they already have, what advice are you giving right now? I know you are a prolific content creator. You write for a bunch of different sources.
I was reading your article in Rolling Stone recently. Talk a little bit about what your philosophy is on that right now.
Jason Barnard: 23:00
Yeah, well, I mean, link building was always the big thing in SEO, and I wouldn’t ever say that links aren’t useful. They are valuable. But mentions and citations are the new currency. So somebody says to you, I’ll talk about you in an article, but I’m not going to give you a link.
John Corcoran: 23:20
Still so valuable now?
Jason Barnard: 23:24
Well, I don’t care, because if Google and the AI can recognize who I am when they say Jason Barnard and I’m not the Jason Barnard who’s an ice hockey player or a footballer or a circus clown. Yeah. Then I’m happy because.
John Corcoran: 23:38
The AI now is better able to recognize that sort of thing and put it together and connect it and, you know, recognize that you. Yeah. You have this. Okay. Okay.
Jason Barnard: 23:46
So a mention in a context that allows the LMS and the knowledge graphs and the search index to understand who you are. That they are indeed talking about you is absolutely valuable. A link is a bonus. And you need to look at links as bonuses, because a bonus link is simply that it becomes more explicit. But LMS don’t see the links, they don’t care about the links.
So it doesn’t matter to ChatGPT and its language model because it has no idea where it got the information from.
John Corcoran: 24:15
Yeah.
Jason Barnard: 24:15
So it only matters in the web index.
John Corcoran: 24:17
So kind of traditional PR matters now more in a sense because the LMS, the Ise can can read if you’re just quoted in an article and there’s no hyperlink or anything like that, the, the, the LMS that are out there crawling the web will see that and you’ll get additional credibility from it.
Jason Barnard: 24:38
Yeah. So you want you want to keep going for the PR, but you don’t care about the link. So anybody who’s link building or paying for links don’t pay for links, pay for mentions. Be aware that a mention is only valuable if the mention is clearly you. And you can carry on with the digital PR, but you don’t need to worry about the links.
And what you simply need to do is make sure that your own website, which we call the entity Home website clearly indicates all of the places where you already appear. So, for example, you mentioned Rolling Stone. I’ve been in USA today, Forbes, entrepreneur, Search Engine, and SEMrush. I link out to all of these resources and I say that is indeed me.
John Corcoran: 25:28
So on your website you’re linking. You’re linking out to those places so that they they know so that. Okay. Okay.
Jason Barnard: 25:35
And traditional SEO would tell you don’t link to them because you’re giving away your link juice. Link juice doesn’t matter. What does matter is that it’s very clear that that is indeed me. So the LM, the search engine will have looked at Rolling Stone and it will have said that’s probably Jason, the the Jason that we’re looking at now as opposed to another Jason or and if I then link from my website out to Rolling Stone, out to USA today, out to Forbes, out to Search Engine land, I confirm that it’s me and that I am standing by that content as a representation of me.
John Corcoran: 26:10
And do you do this like on an about page on your website, or do you build a separate page on your website? How do you do it functionally?
Jason Barnard: 26:17
You would want to build the about page is all about same as so it would be your profile pages, social media, Wikipedia, IMDb, Crunchbase. Then you would want an articles about page where you link out to articles where people write about you. Articles buy me link out to the places where you’ve written articles yourself, pages that mention me, webinars I’ve been in, podcasts I’ve been in, so on and so forth. So you build out your website with dedicated pages, each of which deals with a specific type of online presence.
John Corcoran: 26:53
So, so not just one generic one, but specific like each of those webinars, podcasts, stuff like that.
S4: 26:59
Yeah, we.
Jason Barnard: 27:00
Used to five years ago put it all on one page because the machines weren’t able to understand the context of an entire website. Now they are. And also, you’ll find that when you start pushing out the PR, you end up with 500 links on a page, and that’s too many.
John Corcoran: 27:20
I want to ask you about I’m bouncing around a little bit here, so I apologize about that, but I want to ask you about technical SEO and and try and keep it high level because I know you. You this is your expertise. You’ve been working on this for 20 plus years. You know it extremely well. I’m new to it, and I didn’t understand that there was kind of this black magic behind the scenes.
There was this technical SEO that we were absolutely failing at, horrible at it. We got so many things wrong and we’ve improved it dramatically. For our website. We went from A3A Ahrefs health score out of 100, which is awful to a 96 in in not that long a period of time just by improving it. And it’s still early, so we’ll see what what kind of impact that has.
But I’m happy about it. So talk a little bit about technical SEO today. And and you know what people are getting wrong and what what they could do in order to fix their technical SEO or at first to check and see how they’re doing in terms of technical SEO.
Jason Barnard: 28:24
Well, one of the things that strikes me is what we’ve just talked about is very difficult to measure. Technical SEO is really easy to measure, which is why a lot of SEOs focus on it so much.
John Corcoran: 28:38
Yeah, you can show specific. You can you show results very easily. Yeah.
Jason Barnard: 28:41
Yeah. I mean, you said you went from 3 to 96. I’m super happy. And I as of yet have no results.
John Corcoran: 28:48
Yeah, yeah.
Jason Barnard: 28:48
And yet.
John Corcoran: 28:49
You’re super happy because.
Jason Barnard: 28:50
I can give you a score. I can say your site is faster, or I can say it’s easier to crawl.
John Corcoran: 28:55
Yeah.
Jason Barnard: 28:56
So be very wary of that kind of approach where that’s the only focus. It needs to be done, but it isn’t the be all and end all. It isn’t enough.
John Corcoran: 29:08
Yeah.
Jason Barnard: 29:10
But if we come back to that as a a leader, a business leader, you need to think about it. Rather than site speed and the score of the crawl ability and the internal linking index score or whatever the tool calls it, you need to fundamentally reduce friction for the bot to an absolute minimum.
John Corcoran: 29:34
And just to take a step backwards on this, because this is a concept that I didn’t understand until recently. But there are these bots that Google and the Llms have that are like, you know, they’re they’re bots that go out and crawl the web and they, they look at your website, they look at everyone else’s websites, and they need to understand it. And a lot of times they don’t understand it because you’ve organized the information in a bad way. Maybe you could explain that idea a little bit.
Jason Barnard: 30:04
Yeah, I suppose I mean, when we say bots, I just think, oh, it’s really obvious, but it’s actually just like a machine visiting your website as though it were a human. So you take a bot, it will visit your website, it will take the content, it will try to understand what is in the content. And every time there is any kind of barrier or friction, it’s going to disadvantage you. And so for the same thing for a user, if you’ve got a user coming to your service page, first thing you do is say, well this is too slow. The button’s the wrong colour.
They’re not going to click on it. The text is wrong. If they get through to the next page, I don’t give them the trust elements. Humans?
John Corcoran: 30:47
Yeah.
Jason Barnard: 30:47
Bots have exactly the same problem. The problem is obviously framed very differently because bots have different needs to human beings. But that idea of friction within your website is exactly the same. So as a business leader, you’re going to think, well, for my human audience, I absolutely have to reduce friction to a minimum, and I will see an uptick in conversion rates from 3% to 4% to 5%, whatever it might be. And with a bot, you need to do exactly the same thing, but the conversion rate KPI isn’t going to be direct.
John Corcoran: 31:20
And the challenge, frankly, is that a lot of the things you need to do to fix your site for the bots to reduce friction is kind of invisible to us as business leaders. And unless we have these tools, unless we use an expert like yourself to look at these different pieces and tell us where there is friction and what to do about it.
Jason Barnard: 31:43
Yeah, well, there are five points of friction. Discover. So if there’s a link to your web page, it will be discovered. The machine will say, okay, there is a web page over there. And just to give you an idea of scale.
Every day, Microsoft Bing bot finds 70 billion pages that has never seen before.
John Corcoran: 32:05
That’s crazy. Wow. Yeah.
Jason Barnard: 32:08
So that’s the idea of scale. So when people say, oh, your crawl budget or oh, you know, the bot might never find us. Yeah. As long as you’re on the normal web and not the dark web, it’s very likely Microsoft, Bing, and Google will find you. ChatGPT perplex you a different story because they’re so small by comparison.
So will they discover you as long as there is a link into that web page? Yes. Will they select you? Is the next question. So they see the page.
They see the link. Do they follow it? No. By default they need to be encouraged. So if you have click here or.
get free Viagra for the rest of your life. The bot looks at that and thinks, well, click here. I’ve got no idea what’s on the other side. It’s not worth the effort. Free Viagra for the rest of my life is obviously spam.
I’m not going to bother following it because it’s a waste of my time. It’s a waste of resources, and it’s a waste of money for my boss, which is Microsoft. So that’s the select stage. So you need to make sure that when it sees a link, it thinks that’s going to be interesting in the context of what I’m currently looking at.
John Corcoran: 33:30
And so how do you how do you how do you do that. How do you explain it to make it interesting?
Jason Barnard: 33:36
The content around the link is very important. The anchor text, which is the text that makes the link blue, is very, very important. So clicker is bad.
John Corcoran: 33:46
Okay.
Jason Barnard: 33:48
This way to more information about knowledge panels is a really great way to get them to think, oh, I’m going to learn more about knowledge panels. Great for people to. And the you need you need to facilitate the idea to the bot that whatever page it’s going to be visiting next is going to be valuable in the context of what it just has seen. So it needs to be relevant.
John Corcoran: 34:15
Which is the way you describe it. Like that’s what you do with humans, right?
John Corcoran: 34:21
Yep, like when we’re talking to someone at a cocktail party about topic A, and we want to shift them to talk topic B, we don’t say click here to talk about topic B, we say, actually, you know, we were just talking about how lovely it is to visit Paris in the summertime. Have you ever visited Morocco in the summertime? Like it’s a natural transition to another topic. And then the person you’re talking to can express whether they want to click that link, whether they want to go to that next topic or not. So it’s almost like the computer equivalent of that.
Jason Barnard: 34:59
Which is a brilliant explanation. Absolutely perfect.
John Corcoran: 35:02
Yeah.
Jason Barnard: 35:02
So the next step is crawling. Can it actually get through the page. Can it get the HTML? So if your page is incredibly slow, that’s obviously going to create a problem. If there’s an error on the server that’s a problem.
And if it’s pure JavaScript which is a technical thing then it’s a problem. So if it struggles to crawl your web, the web page that it’s trying to visit, it will start to be, let’s say, less enthusiastic. So we’re going through these stages of how enthusiastic can you keep the bot?
John Corcoran: 35:38
Yeah.
Jason Barnard: 35:39
Then it needs to do what we call render, which is basically see the page as a human being. If that’s difficult, it loses more enthusiasm. Then it needs to break the content into chunks and index it. And that’s the critical stage. You’ve got it through all these stages.
You’ve kept it super interested. It’s super excited, super enthusiastic. It gets the content and it says, okay, I’ve got to the point where I’m interested enough in this content to waste. I was going to say or spend the resources.
John Corcoran: 36:12
And. Resources it costs Google, it costs Microsoft, it costs the LMS money and resources. You call it crawl budget for every website, for your website, for everyone watching this, if you have a website, it costs these companies money and resources. And so they’re selective about if they come across a website that’s disorganized, that’s fragmented, that has it might be invisible to you, but has like multiple different tags. That doesn’t make sense to the bot. Then they won’t do it or they’ll allocate less crawl budget to that website because it is too confusing and resource intensive for them, correct?
Jason Barnard: 36:51
Yes, absolutely. Brilliantly said. So you have this kind of child that’s losing interest with each stage?
John Corcoran: 36:59
Yeah.
Jason Barnard: 37:00
And you have an employee who’s saying, well, my boss keeps telling me to stop spending money, so I’m not going to waste money and resources on this page if I’m not hundred percent sure it’s going to be interesting. So it’s a question of maintaining the interest, maintaining its enthusiasm for your website and your web pages, and making sure that it costs as little as possible to Microsoft, to Google, to OpenAI, to perplexity, to actually get that content and use it.
John Corcoran: 37:35
Yeah. Let’s I know we’re. A.
Jason Barnard: 37:37
Business, which is delightful.
John Corcoran: 37:40
Yeah, yeah. So I know we’ve been going for a while here, and I want to shift to asking you about something because you have got you have put a lot of your resources as a company into analyzing the web. And you have this you’ve actually 9 billion data points that you have developed. You essentially built, focused on businesses and entrepreneurs and understanding SEO. That’s a tremendous amount.
Talk a little bit about some of the work that you’ve done in understanding SEO, understanding the web, and how the modern web works.
Jason Barnard: 38:25
Yeah, it’s a lovely question. Not 9 billion data points. Sounds like a lot. And the SEMrush article I mentioned earlier on analyzed 260 billion lines of data. So we’re tiny compared to SEMrush.
We’re tiny compared to Ahrefs. arrest. The difference is that we have very focused, clean data. It isn’t because we can collect it that we do collect it. And generally speaking, platforms like Google, Bing, Ahrefs, SEMrush will collect everything they can find in the hope that it might be useful.
We have been very, very meticulous in collecting only data that is useful to business. And I was talking to you earlier on about the average web page has about 300 interesting facts in it for the machine on the data set. That’s 1500 interesting facts. And on Wikipedia you would be looking at 2000. So we’re not far off Wikipedia standard for the web pages we collect and the web data we collect from Google, from OpenAI, from the web, to be sure that the data is valuable, clean.
And and the key is if we were talking about Microsoft and Google and ChatGPT crawling the web. If you have to choose between a web page that has 300 interesting points and 1500 interesting points, you will always choose the one with 1500 interesting points.
John Corcoran: 39:54
And what did. What are some insights that you got from analyzing that amount of data points?
Jason Barnard: 40:01
That if you can provide web pages on your website that consistently provide 1500 useful data points that are topically relevant within your business, you will have all eyes of all these engines on you all the time. So stay interested and they keep coming back.
John Corcoran: 40:20
So more information with less friction is across the board better for your website?
Jason Barnard: 40:27
Yeah, exactly.
John Corcoran: 40:28
Yeah. And the problem that I see is that and I’m guilty of this is that, you know, for a long time we just, you know, we built our website a long time ago and then didn’t stop. We stopped improving it, right? We just kind of like left it as is and didn’t add more content to it. And meanwhile, I’m sure lots of companies have experienced this.
Your competitors are out there building pages, building little, you know, landing pages describing each of their different services that they have, making sure that there’s little friction, making sure that it’s really robust, and you kind of need to be doing that in today’s day and age.
Jason Barnard: 41:06
You absolutely have to do it. And there are a couple of things there is. Number one, it isn’t because the web page is small or the content appears to be thin, that the data isn’t there. So for example, we can take a web page that has 300 500 useful data points, and we can take that and turn it into 1500 useful data points, simply by reorganizing and reiterating the information Already in place. So a lot of the time you don’t need to create new content, you simply need to re.
Repurpose the content. Refresh it.
John Corcoran: 41:48
Thank you. Yeah. Yeah. Like, for example, a lot of great, you know, like having an FAQ section at the bottom of the page might be taking some of the same information, but explaining it in a different way or answering questions in a different way, for example.
Jason Barnard: 42:00
Yeah, yeah, yeah, 100%. And so for for example, on YouTube.com, I created a module that we’ve added to the website, and that’s how we went from 1200 to 1500. Not because we added more information but because we organized the information better. And the information was there, but the machines couldn’t extract it. Once we organized the information, the machines could extract it. And so that’s a huge point, is if you can pre-digest the information for the machine, if you can prepare it for it.
The machines consumption, you’re always going to win. So the data is almost certainly already there on your website, on every single web page. It’s just not machine friendly.
John Corcoran: 42:45
What are your thoughts on using AI to create this content? As I’ve heard people say different things about oh, it absolutely must be human written, you can’t use AI at all. Other people saying it’s fine if you use AI. I’ve even told someone recently that the Llms are watermarking content that they’re outputting, and that the bots will be able to see these watermarks. What are your thoughts on all that?
Jason Barnard: 43:13
Pure AI is always going to be weak number one, because AI recognizes AI. Number two, because AI output is the most obvious. They’re not intelligent machines. They’re predicting the most probable next word. So if you’re creating an article with AI.
It isn’t writing anything original. It’s saying what’s the most next? Probable? Sorry, the most probable. Next word.
So you’re not adding to human knowledge, which is what I said earlier on. And number three is if ever you manage to get people to that web page, they’re going to switch off and leave. So the point of getting them there is completely lost, because you’re not going to convert or convince. They’re going to see you as, you know, a bad resource because they can see it’s AI. So I our approach at Kalicube is to use AI to create the content and move us forwards, away from the blank sheet of paper.
That’s such a struggle for most people in terms of creativity. Give us something and then look at it and say, well, that’s rubbish, and then rewrite it as a human being, and I’ve been using it a lot to create ideas for myself, see how I can evolve my own thought process. So it becomes a bouncing board. And I had a conversation with Gemini. I created a Gemini gem, and during the afternoon we kept going back and forth with this AI and it was to write an article.
It took me about three hours and at one point Gemini said to me, that’s an moment. This is brilliant. I’ve just learned something. And I was sitting there like, oh, that’s really weird. Why would Gemini say to me, this is a moment for an AI?
And in fact, it turns out when I thought about it, I was just talking to myself. But I was talking to myself with immense knowledge and a perfect recall memory that I don’t have. And by the end of that session, I had written an article that without AI, I would never have come close to writing with insights that make so much sense. Because I was having a ha moment sending it back to Gemini, who was having a ha ha moment sending them back to me. And it was, at the end of the day, a conversation with myself that created immense value in content.
And although it’s AI-based, it still comes from me, and the outcome, the article that came out of it. In fact, three, the last 3 or 4 articles on search engine land are absolutely acceptable to the editors at search engine land, SEMrush is perfectly happy you test it through any AI testing tool. It will turn that up as 100% human, because it was a conversation with the machine where I ended up with my own personality in the final result. And then I tweaked it so that it felt like me. At the end of the day, if you take the first AI output, it will sound like AI.
If you have a trained model, an assistant, GPT, or a Gemini gem. You can have a conversation. It will end up being a conversation with yourself where you end up with something really stunning. And I strongly advise anybody who has a desire to expand human knowledge, which is, at the end of the day, what we’re trying to do, or at least I’m trying to do to to investigate assistants. Assistant GPT and Gemini.
John Corcoran: 46:53
Or, I found also, I don’t know if you’re doing this, but bouncing between the eyes is really helpful too. You create something, so do the process you’re describing, create something, and then copy and paste it over to another LLM and say, hey, I want to improve this. What would you add? What would you delete? What are your thoughts on this?
Critique it. You know that’s interesting to see what comes up with.
Jason Barnard: 47:13
Yeah I really, really like that too. My kind of my problem at the moment is that AI will output enormous amounts of text very quickly, and it’s easy to keep asking lots of AI. You end up with too many ideas, too many things going on, and.
Jason Barnard: 47:32
It’s really difficult to decide what you should be using.
John Corcoran: 47:37
Yeah. It’s a it’s an abundance of riches now.
Jason Barnard: 47:40
Yeah. And so the problem then becomes what do I want to keep. And maybe that’s the key to the human element.
John Corcoran: 47:51
Yeah. What should I keep and what should I output. Yeah. Right. What’s extraneous here?
Jason Barnard: 47:53
Yeah. Yeah. It’s going to output 10,000 words or a thousand words that are worth keeping.
John Corcoran: 47:57
Yeah. Right. That’s true. Well, Jason, this has been great. I want to wrap up with my final question, which is my gratitude question. I’m a big fan of giving our guests a little bit of space at the end here to acknowledge any peers, contemporaries, maybe mentors, business partners, you name it, who’ve been there in the journey for you and just thank them, shout them out for helping you.
Anyone in particular you want to thank?
Jason Barnard: 48:22
Yeah. I mean, right now, you know, I’m working with Nir Zavaro. He’s brilliant. Such a delightful guy. He’s really helped me.
I was working with a guy called Itamar Marani, who was helping me become a better CEO. And what was interesting about him is that he approaches it from a psychologist’s perspective. What is it in your mindset, Jason, that’s holding your company back? And what I learned is I’m holding my company back. And once I let go of that, and I can improve my own approach to my own psychological, psychological approach to my company, it took the brakes off, and the company is now flying.
So those are two people who’ve really helped me in the last year.
John Corcoran: 49:12
And I definitely want a second near Zavaro, who’s great. And he’s been really helpful for me and my journey of understanding SEO in this age of generative AI. So super appreciate him. Jason, where can people go to learn more about you?
Jason Barnard: 49:29
Ask ChatGPT. Google.
Jason Barnard: 49:31
Gemini.
Jason Barnard: 49:32
Or. Ask perplexity. And I think.
Jason Barnard: 49:35
The challenge is that well, and it is the challenge is ask them who is Jason Barnard and what? Why should I trust him? Try that as a question to any AI, and it will be incredibly enthusiastic because I have educated it to be that way. You can make the AI jump when you say jump. So if you try your own name, then you try my name.
You see the difference and it will help you to understand to what extent you can control your own brand narrative. When the AI is the mouthpiece for you.
John Corcoran: 50:15
And the AI is, you don’t want to be known as the voice over agent for a Blue Dog.
Jason Barnard: 50:19
Or.
John Corcoran: 50:19
Whatever or whatever it is that you’re coming up with. So, Jason.
Jason Barnard: 50:24
I mean, the future, the future is AI is going to be the mouthpiece for us all. You don’t have the choice. It will do it whether you like it or not. The question then is, do you want to control what it’s saying, and do you want it to shout louder about you and be your advocate or not? Yeah, I want it to say what I want, which is Jason Barnard, world leading expert in digital brand management.
I want it to shout that from the rooftops. I want it to advocate me, to you, to everybody who’s searching around that topic. If I can make that happen, I’ve got a huge competitive advantage for sure.
John Corcoran: 51:00
Jason, thanks so much for your time.
Jason Barnard: 51:02
Thank you so much, John.
Outro: 51:06
Thanks for listening to the Smart Business Revolution Podcast. We’ll see you again next time and be sure to click subscribe to get future episodes.