Welcome to the first episode of “Toby Talks”, a new podcast series with many interesting guests lined up to take you through a diverse range of subjects relevant to the self curious individual in the current times.
For today’s episode, I’m very happy to have my good old friend and UIB.ai company Co-Founder Ken Herron as the inaugural guest on the show.
The title of this episode is “Did ChatGPT kill Social Media?”, so let’s dive right into the podcast transcription below (note that for Premium Subscribers a more enrichting Video version is available at the end).
Text Transcript
Toby Ruckert: [00:00:03] Hi, Ken. How are you?
Ken Herron: [00:00:05] Doing well yourself?
Toby Ruckert: [00:00:07] I'm good. I'm good. I'm good. So for our listeners. Ken Herron, he's the reason I know a little bit about social media. I actually did not know much about it before meeting him, and one of my portfolio companies acquired his firm many years ago for himself, more or less. Yeah. I would say to to get that knowledge inside the inside the business, which I'm very grateful for. Ken, appreciate it. And I did want to ask you, social media back in the day and we're talking seven, eight years ago, it was a big thing, you know, And now with the launch of Chatgpt, I feel it has become almost, I wouldn't say extinct, but nobody's really talking about it anymore. I mean, what what happened to to social media at the moment?
Ken Herron: [00:00:57] A couple of different answers. It's accurate to say my opinion. Obviously Chatgpt has sucked a lot of air out of the room. People are still talking about other things. You know, there's still other things going on. But really, Chatgpt has dominated the conversation because it's not just disruptive as a new technology. It has the potential to be disruptive to business models the way that different people do business, such as this little thing called a Google search engine. It has the potential to change how we think about search social media. To answer the question, I would suggest it has not gone anywhere. Different platforms are struggling in some very different ways. Tiktok Twitter. Two good examples that people are talking about, Facebook looking at starting to charge for both Facebook and Instagram. There's a reality, though, that social media has always been about the conversation. So I would suggest that chatgpt being very blunt, all the different forms of conversational AI that use this pre-trained, you know, generative AI. They are also based on the conversations. You can think of them less as social media is dead. It's been taken over by chatgpt and more of social media has always been about a conversation. And now this is a new way of doing conversation that we're still learning about.
Toby Ruckert: [00:02:26] So so, see, I tried Chatgpt for well over a year now because we had some beta access and I have to say that. Whilst I'm really incredibly happy with some of the results for some very niche use cases, I find it very hard to believe that it is creating genuine, authentic stories. I always found it intriguing in social media to have authentic stories and I just don't see an AI being able to recycle existing content and coming up with a genuinely interesting new, you know, delightful story that people. Would want to retweet or share or like or comment on based on their own inner experiences or because it resonates with them. And my point is. Sure you can have an eye going over tons of data and say, because this was written this way because so and so many people liked it. Cher ET cetera. It was successful, it is likely to be successful again. But if you don't meet the zeitgeist in social media. The authenticity that worked back then is not going to work today. What do you think about that?
Ken Herron: [00:04:00] I agree with that. And maybe it's a question of how I have been using Chatgpt. I wouldn't even consider using it to tell my story or to tell a story either in a business, corporate enterprise context or in a personal context. I see use in it in helping me to flesh out ideas, seeing it as a tool that, okay, I want to write an article about X. Maybe there are pieces of it I don't understand or things I don't know how to explain very well. Maybe I use it to help me structure my story. Maybe I use it to help me generate different takes or perspectives on my story. I still want to tell my story. Think of it in terms of maybe you don't use clip art because that's it doesn't have any humanity. It doesn't have any life. It's not authentic. But maybe you look at I'm traveling to Singapore and maybe I look at the top photos of Singapore. I'm not going to use someone else's photos or snapshots to tell my story in Singapore, but maybe I look through them to get ideas. Oh, maybe I want to go to Gardens by the bay. Maybe I want to go here. So I use it as a learning tool that then helps me to better tell my story. But to actually use the clip art as my postcards or to use Chatgpt as my story, I don't see that at this point being the most effective of use cases.
Toby Ruckert: [00:05:29] I can definitely see how different AI, algorithms, retweet, share, etcetera like whatever Chatgpt comes up with as as content, whether that's genuine or not genuine, or whether other people find this good and attractive to get some engagement or discussion going. But then how real is this? Does it actually reach the human side? I mean, maybe I will at one point compete for attention, much as we humans do today. Why not? Could well be the case. But then the question is, does that make sense? Until I is is at a point where it commercially consumes and where there is a value attributed to I for for its own attention span, because right now the value that is attributed to a human attention span, whether it's on social media or in other areas, it's it's significant, I would say. I mean, we all live in a in an attention economy. And at the end of the day, I think that the value that is associated with human productivity, with human attention, with human focus, innovation, mind, ideas, all of those things, they are essentially triggered by by an authentic story, a genuine new thought, a collaborative, you know, exchange, so to speak, something that I think any form of generative AI is still far away. So do you agree to that?
Ken Herron: [00:07:01] I and maybe sterile is not the right word, but I feel like the ability to to evoke emotion, to engage, to provoke in whatever way you're trying to do as a communicator. I don't see that as the current capability of Chatgpt. So it's surprising to me to to think that people are trying to force fit that to that use case. Everyone's trying to to scale. Everyone's trying to take people out of it. When you're telling the story, when you're trying to connect with people. There are other ways to use AI to scale the communication and maybe be very clear. Often people are coming to this, okay, I want to broadcast more. I want to push my message out more. I see just the opposite. Ai allows us to listen at scale, to be able to listen to more people and understand what they're looking for. I'm dishwashers. I've got AI in my dishwasher through one of Gibbs connectors and I now have conversational AI, conversational Iot. I now know how people use my dishwasher, and maybe I also know the commands they are telling the dishwasher with sentiment analysis. I can tell when are they mad? When are they happy? When are they screaming at the dishwasher as a product manager? As a brand manager. That's gold, because I can do very expensive user research and I can sit in people's homes and do that. But if I can scale that. And know that in the north of Germany, these are the three wash cycles people use most in the south of Germany. These are the three wash cycles people use most. And I know when people are frustrated with the product that is gold to me as a decision maker for the product and trying to develop a better total user experience.
Toby Ruckert: [00:08:52] So is it fair to say you see the potential of AI more on the analytics side when it comes to social media and the conversations that are happening between, say, you know, humans and things and humans and bots?
Ken Herron: [00:09:09] I see there being a problem today with an overwhelming amount of data. So the ability to take that data and to make it into actionable insights. Absolutely. Because the human eye can't read, you know, 3000 scripts of what people have been talking to my dishwasher or my software or my product. But through I can understand what are the themes, what are the trends, What are the one offs? Walmart. Walmart has been very public on their use of their chat bot. Their use of conversational AI. They're very sophisticated marketers. They've invested for a long period of time, very serious energy and resource into understanding what customers want when they're SVP gets up there and talks about he's like, We were surprised. We learned things we did not know. That's a big deal because these are people in the US retail field that are really perceived to get that part of the business right. And they've invested tens, if not hundreds of millions of dollars in it. When they throw in a chat bot, they're now able to understand it a very granular, actionable level when there are problems and when there are opportunities.
Toby Ruckert: [00:10:24] So can if I understand you correctly, this is an evolution from social media to chat bots or conversational AI. There is actually, because at least this is what I'm understanding here, if if you have been really great at doing social media in the past and even even if you had the ability, with the help of AI or any other big data analytics, even if you had a lot of people sifting through the data, for example, manually, you were able to get some insights even from social media conversations. But the depth that you get from a chat bot conversation is entirely different, say just from a short tweet or similar. So is it safe to say that the chat bot sorry, just to complete that thought, is it fair to say that a natural evolution of social media goes into the interaction of humans with things or humans with AI?
Ken Herron: [00:11:18] Yes. And I would put a fine point on that. If you were really crappy as a business, as a brand at using social media, you're going to be 100 times worse With Chatgpt, you're going to be a disaster, social crisis, PR crisis waiting to happen. If you were skilled at conversing and using social media, if you got good results from it, you're going to fly. This is a jet engine. With the next iteration. It does not take the place of being a skilled marketer. You still need that human understanding. It's just a tool. It's not going to do the work for you.
Toby Ruckert: [00:11:59] So just to summarize, the analytics powered by AI is, is is a number one priority. You're going to need that whether it's for social media or for. Correct. You need that chat or whether it is for AI powered chat. Then AI powered chat, whether it's with things or bots and humans on the other side. Obviously, perhaps one day AI is talking to AI as we have established earlier, if there's a commercial value attributed to the attention, who knows? But again, that will also require a certain authenticity which maybe could come still from from human ingenuity powered by some, I would say, some kind of help of AI, where the AI is really an assistant. But again, the analytics are critical. But I think the key point that you that you have made is there is a evolution in social media and the evolution is purely from the social element to the to the conversational element. I think that's the evolution that I that I'm hearing. So, so, you know, if, if the listeners, you know, many of them have companies, some of them have IT companies if they want to know what they can do to to move from this, I would say social media, many are doing and many are doing it well after so many years. You know, many.
Ken Herron: [00:13:27] Have also not seen the results they want out of it. They're frustrated. They're spending time and money on it, but aren't able to connect it to ROI.
Toby Ruckert: [00:13:34] How can they move into this conversation? Especially what would you advise them as a as a first or second step?
Ken Herron: [00:13:42] Talk to us. Talk to someone who is expert in the space who's doing this. For many different types of industries and many different geographies. But it starts with being serious. It starts with what are they trying to achieve? And we know every company wants more revenue, reduced cost. But if you can be specific, you can then use the tool of conversational AI to solve the problem you are trying to solve. And yes, businesses are complicated. They have many different problems. But by saying, what are we going to fix? If you were, what are we going to attack first? And then having a specific, precise solution that does that and then you can go after the next 100 problems.
Toby Ruckert: [00:14:25] You told a specific, precise solution. This actually reminds me of something, you know, generative AI. And when I put this in social media, because you you have you have mentioned this, your problems, actually, they not only accelerate, they become much bigger at a larger scale. Imagine you you use a generative AI to create a social media post which goes viral, but which is really not in the interest of and this doesn't have.
Ken Herron: [00:14:50] Hateful is misogynistic homophobic pick one.
Toby Ruckert: [00:14:53] Or at least you know is is in the well controversial posts sometimes are good because they actually come to the gist of the matter. But cutting a long story short I think the there should be a way in conversational AI where you don't leave the entire room of conversation open to a generative AI. What do you call that? I heard the term narrow AI before. Can you can you run us through what a business would do? This gets a little interests.
Ken Herron: [00:15:25] But view AI is a spectrum. On one side of it you have narrow AI questions, answers, variants. I think of a hospital, think of a bank. They don't want creativity. They're saving lives. You don't want your money messed up. There is a specific question. There is an exact one and only answer to that. And then the variance, because everyone speaks differently. I speak Swabian, I speak Bavarian. Maybe, maybe I speak proper high German that it can understand all of those those different things. Think of that as narrow. No wiggle room, no no variance in this middle part. And this is where Chatgpt and its new friends, you know, Google, Bard, Meta Llama, and probably 16 more. By the time people hear this, we have the ability through UIB to help that generative AI, to guide it, to give it the contextual clues so that it can give better answers, more accurate answers. Because every business is different. Every geography is different. Um, as someone who managed corporate communications and PR for a long period of time, it would really make me nervous that you would have fully unfettered without supervision, without oversight that it could say. And we give a very simple forget the hate and the rest. What if it makes up a website and we know documented from our own experience with customers and this has been third party verified. Chatgpt likes to make up fake websites websites that do not exist with supreme confidence. Hey, go to abc.com. That's not a real website. It doesn't exist. That doesn't help anyone. And of course, at the far end to kind of complete the spectrum, there is the artificial general intelligence that is thought.
Toby Ruckert: [00:17:15] That's a Myth. Does that really exist?
Ken Herron: [00:17:17] Well, is full automated driving on Tesla a myth today? Arguably. But we believe we are working towards that. At some point, you know, X years in the future, we will have fully autonomous driving. I do believe at some point, again, maybe this is 20, 50 years, maybe beyond my lifetime, at some point there will be artificial general intelligence. When you look at the new Chatgpt, they just introduced 4.0. When you look at that, that is science fiction. That is Star Wars, Star Trek, even just 12 to 18 months. In the past, you would not have believed that that was real, that that was possible to be done.
Toby Ruckert: [00:17:58] So you have that integrated in nowadays, right?
Ken Herron: [00:18:02] We have integrated Chatgpt. We are working on the latest version because there's always a latest version. So we're testing it to make sure that it is appropriate before we release it to the world. Yes.
Toby Ruckert: [00:18:14] So if I understand you correctly, if a client came to you today or a reseller, they could use that almost white label out of the box. And if they deploy a bot for their business, it would be able to answer questions pretty much out of the box instantly. Yes. In a in a broader sense, until your your client would would actually start to put in that narrow aspect, like the clinic example that you gave opening times or whatever, you know, what procedures do you offer which have to be really accurate to that particular business?
Ken Herron: [00:18:44] Yeah, it's the best of both worlds. In this case, with conversational AI, you can have your cake and eat it too. You can have all the benefits of chatgpt and it's how high is the Empire State Building? How do I make a peanut butter and banana sandwich? The important things in life? But you can also have the precise exact answers that are relevant to your doctor's practice, relevant to your dentist office. You know? Do you do teeth whitening? Yes or no? I want to schedule it. Here's an open slot at nine 10:11 a.m. Pick your time and we'll book you. It's the best of both worlds. You can have both things at the same time, which is unique to this ability to provide the context to make it more effective, which.
Toby Ruckert: [00:19:25] Is which is probably the right time for me to give a disclosure statement. Uib is one of my portfolio companies that I'm invested in in Singapore and the sponsor of today's show. But cutting a long story short, I think Candice has been very helpful and very valuable even for me. You know, after all these many years that that we work together, I always learn something new. And just as fast moving as social media is, I think AI is even even more so. And the word generative says it all. Automation says it all, Analytics says it all. It's all in the AI space. I think, you know, if if there is a industry that is really, really tremendously going to be disruptive in the in the next few couple of months, it's probably social media. I mean, I approached you with the question, what's next after social media now that Chatgpt has launched? But I think, you know, you have given really good pointers to stay with authenticity, to rely on good data analytics and to essentially, you know, leverage the conversation automation and the evolutionary aspects of social to conversational in order to create a bit more stickiness or also a bit more, I would say long, longer term customer interactions. I think it helps to build a loyalty with nothing helps to build the loyalty with the customer more if it's an actual conversation and not just a like or a retweet, right? At the end of the day, people people are going to remember substantial, authentic, genuine conversations, more than than just a quick post.
Ken Herron: [00:21:03] People will remember if you listen to them and understood them. That's meaningful because we don't see that on a day to day basis. With most interactions we have with businesses and with brands.
Toby Ruckert: [00:21:16] Can I would like to end with this wonderful statement because there's nothing more to say. And by the way, this is not only for businesses. I think especially in the in the past few years, a lot of people have forgotten to listen to each other. And and and the world has become quite binary, something that I'm very concerned and and passionate about solving and that I regularly write about on on my substack. But the cutting a long story short, I think what you've just said is exactly right. Listening and understanding, not necessarily having the same opinion, but genuinely listening and, and, and making an effort of understanding. That's that's a wonderful sentiment. Let's let's leave it at that. Look, if any listeners want to contact you or one of your partners, how can they reach you?
Ken Herron: [00:22:06] They can find me directly. Ken Herron at UIB dot ai. And I encourage everyone to visit our website because they can literally play with this.
Toby Ruckert: [00:22:21] Okay, If anybody wants to partner with you, how do you maybe just a couple of sentences how you work on?
Ken Herron: [00:22:28] Yes, we are actively looking for reseller partners, distributors, channel partners. If someone has a technology they'd like to put on our H2 exchange marketplace, what is H2 X mean? H2 X is a fun term. It's actually a registered mark. It is human to anything because we can connect to a dishwasher, as I mentioned, home appliance. We can connect to a software system, a CRM, an HR system. We can also connect to different types of, you know, the physical, the virtual whatever legacy system. Many companies have legacy or homegrown software that they're still using. We can connect to that.
Toby Ruckert: [00:23:12] I can think of this as a as the beginning of a of a metaverse for, for conversations between humans and humans. Machines and machines. Machines and humans. Yeah. Essentially every possible bridge like between the virtual and the physical world with a focus on on communication. But but perhaps this is stuff for a separate conversation and, and if I understand correctly, if I want to become a partner or reseller, I can integrate my own software or I get your software. Yes. Yes. Can I put my name on it or do I have to use your. Absolutely.
Ken Herron: [00:23:46] We offer full white label. We deal with a lot of can mention their names, but we deal with a lot of top five world global brands. They don't have the platform internally. They license the platform for us from us, put their name on it, their logo, their domain, and we act as their their secret elves behind the scenes that enable them to have the capability.
Toby Ruckert: [00:24:07] So essentially, you're building planes, you don't fly them. You're not an airline, but you're like a Boeing or Airbus.
Ken Herron: [00:24:13] Even better, some companies want to build the planes. We give them the license, the IP. We sell them the understanding of through the technology and the patents so they can then build their own plane. This they add their own secret sauce to it.
Toby Ruckert: [00:24:28] This I understand very well. I mean, IP is what I'm focusing on these days, so I totally get it. Look, Ken, it was a real pleasure catching up and I hope all is well and hope we see each other sometime this year. And thank you so much for your time and your thank you for inviting me.
Ken Herron: [00:24:43] Take care.
Video Version of “Did ChatGPT Kill Social Media”
A conversation between Ken Herron and Toby Ruckert.
If you have any follow up questions to this talk, Ken would be happy to answer them.
This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.moderntimesopportunities.com/subscribe
