Artwork for podcast Creatives With AI
E43 - Mastering Innovation and Entrepreneurship in the AI Era: A Deep Dive with Sophia Matveeva
Episode 4315th March 2024 • Creatives With AI • Futurehand Media
00:00:00 00:55:45

Share Episode


In this episode, Sophia Matveeva discusses the shift from perfectionism to failure in entrepreneurship and the impact of the education system on preparing individuals for the digital age. She explores the evolution of digital products and the rise of non-technical founders in the tech industry.

Sophia also highlights the democratisation of tech innovation and the convergence of AI and business. She emphasises the importance of understanding the three stages of technological adoption and the challenges of hardware and robotics. Finally, she encourages questioning data and considering the unintended consequences of AI.

The conversation covers a wide range of topics related to AI and technology. It explores the unintended consequences of social media, the importance of learning AI theory, and the value of hands-on experience with AI. The discussion also delves into using Chat GPT as an annoying assistant, the evolving functionality of Chat GPT, and its usefulness for summarising information and analysing websites. The conversation highlights the importance of clear brand messaging and references Red Bull's successful branding strategy. It concludes with a discussion on the importance of politeness in interacting with AI and promotes a free ebook on technology concepts for business leaders.


  • The shift from perfectionism to embracing failure is crucial in entrepreneurship.
  • The education system needs to adapt to prepare individuals for the digital age.
  • Digital products are never complete and require continuous iteration.
  • The rise of non-technical founders in the tech industry is a promising trend.
  • AI is democratising tech innovation and changing the landscape of business.
  • Understanding the stages of technological adoption is essential for success.
  • Hardware and robotics present unique challenges in the tech industry.
  • People play a critical role in the development and use of AI and robotics.
  • Questioning data and considering unintended consequences is important in AI development. Social media has had unintended consequences, including psychological problems in children.
  • Learning AI theory is important for understanding and effectively using AI tools.
  • Hands-on experience with AI, such as using Chat GPT, can be a valuable learning exercise.
  • Chat GPT can be used as an annoying assistant, providing initial prompts and sparking creativity.
  • The functionality of Chat GPT has evolved, with some limitations and adjustments.
  • Chat GPT can be useful for summarising information and analysing websites.
  • Clear brand messaging is important for effective communication.
  • Red Bull's branding strategy focuses on lifestyle and subtle logo placement.
  • Politeness in interacting with AI is important for shaping human behaviour and training AI.
  • A free ebook on technology concepts for business leaders is available for further learning.

Links relevant to this episode:

Thanks for listening, and stay curious!



Tools we use and recommend:

Riverside FM - Our remote recording platform

Music Radio Creative - Our voiceover and audio engineering partner

Podcastpage - Podcast website hosting where we got started


00:01 - David Brown (Host)

Hello, welcome to the Creatives WithAI podcast. I'm your host, David. On today's show we have Sophia Matveeva. This is another hard hitter. I've got really, really experienced fantastic people on early this year and she is another one. Sophia is the CEO and Founder of Tech for Non-Techies. It's an executive education and consulting company. Sophia also has contributed to the Harvard Business Review, Financial Times, the Guardian, Forbes. She writes on entrepreneurship technology and she also hosts the most excellent and top rated Tech for Non-Techies podcast, which has about 100 and something episodes. I looked at that today, so no pressure.

00:43 - Sophia Matveeva (Guest)

It's almost 200.

00:44 - David Brown (Host)

She's also a guest lecturer at the University of Chicago, London Business School and Oxford University. She's a startup mentor at the Chicago Booth Polsky Center for Entrepreneurship. She's advised on leading accelerators, including the Chicago Booth's new venture challenge and Techstars and Blackstone Launchpad, which I'm sure she'll probably talk about some of that in a minute. She's also got an MBA from Chicago Booth and a bachelor's in politics from Bristol. She speaks multiple languages, including English, Russian and French, and she sits on the advisory board of Riveter, which uses AI to predict consumer trends for the world's biggest brands. I'm sure there's more to this than we're not even covering, but I'm going to stop there and just introduce Sophia. Sophia, Sophia, welcome to the show.

01:31 - Sophia Matveeva (Guest)

Thank you so much, David. I am so happy to be here and I'm excited about this conversation. You mentioned the Tech for Non-Techies podcast and the number of shows. What I want to say to all listeners is that we always have drama about getting started with something, but then, once you are on the path to doing it and it becomes a habit, then you just get on with it, because the biggest drama I had was about starting the podcast. I wish that I had started it earlier because I had the material.


I was writing for Forbes. I could have just literally recorded those interviews and had a great show, but I was having all this drama about will I have time, will I do it? It's such hard work, blah, blah, blah. But, David, as you know, when you start producing a show and it has to come out at a regular cadence, it just becomes part of your identity and you just get it done and it will be kind of like that's now what I do. So, to everybody, whatever it is you want to do, you get started. Make it a habit. That's how you get things done. The biggest drama is at the beginning.

02:40 - David Brown (Host)

You're absolutely correct. What's been really interesting for me is to watch how because I was born in the late sixties and so I was raised in an entirely different world and one of the things that always was kind of taught to us is, when I was young, it was bad to fail, like you were supposed to be successful all the time and so if you weren't going to be successful at something, then you learned that you just didn't do it really, whereas now there's a huge, the whole concept of everything, pretty much startups and business, and everything is fail, fail fast, it doesn't matter. If you fail, you've got to try, and you're absolutely right, and I think it's taken me. I mean, I've had to go on almost like a personal journey, which sounds like it's slightly dramatic, but I've had to go on this personal journey to get to the point where I can just be biassed towards action and I try all sorts of things because not all of them are going to work, but you've just gotta keep trying and you've got to keep doing it.


So, yeah, I totally agree, but yeah, that's how it was back then when you were in school. It wasn't really exploring and testing new things and that sort of thing Like if you were in science class you could do an experiment, but if you were outside of that experimenting and that sort of thing really wasn't, it just wasn't a thing that people did, and I think that's why there's a whole group of us kind of older I hate saying that middle age people and up who don't really it doesn't sit comfortable with us in a lot of ways, whereas a lot of the younger you know my son is quite happy to experiment and to try something and to fail and doing that fail fast thing and just see if something works and if it doesn't. So yeah, it's great advice.

04:36 - Sophia Matveeva (Guest)

You know, I think it's also. It's not just the education system, it's also the way it's a hang up from the industrial revolution and the information age and where we are now basically requires a different skill set. So I'll give you an example. So right now you're drinking from a cup, and so that cup had to be made for you in a factory. So what happened is that somebody decided okay, there is a market for this yellow cup, so we're going to make these cups. Then they got some designers. A designer designed the cup, then they gave it to manufacturers, they manufactured it, and then it was marketing and sales. So I don't know whether it was a website or Amazon or whatever some sort of marketing itself. The aim of marketing and sales is, you know, to find people like you who are going to buy this cup, and then, once you make the payment as a customer, the cup is sent to you. The company's life with the cup or the product it ends, and then your life with the cup begins, and so there is no overlap. The only overlap that you as a customer have with the company is at the time of transaction, and the time of transaction and therefore the cup, the final product, it has to be perfect. You know, if it was sold to you and it was a bit sharp and like, it wasn't kind of smooth and it cut your lips, you wouldn't be happy with it. Or if it had cracks. Or you know, if it was kind of half painted, you wouldn't be happy with an imperfect product. And so, before the digital age, so all the products that are basically not digital products, they have to be perfect because they have to be complete by the time the customer gets them. So that's true about the clothes that I'm assuming most of the listeners are wearing right now. It's also true of, like, legal agreements. You know, for example, I mean, god forbid, god forbid, you're getting divorced. And then you have to go to a divorce lawyer and the divorce lawyer says to you well, you know, here's your divorce contract. I think you're divorced, but I'm not entirely sure. So why don't you try getting married again and let's just see what happens? Like, do you think that person was crazy? And you would definitely not recommend them to your friends, right? And so this, the education system, is designed to prepare people to be employable at the end of the day. But it's not the only point, but it's one of the most important points, and so the education system has been preparing people to be employees that make perfect products, whether it's a cup, clothes, a building, a legal contract, whatever it is.


In the digital age, products are never complete and therefore they're never perfect. And what I mean by that is look at the apps on your phone. They change without you doing anything. So you know how, like a logo, like the Uber logo, can change, and then sometimes it's hard to find because, oh, you see black and white and now it's powerful, and you're like what am I doing?


Or the screens in an Apple change, which is sometimes great but sometimes annoying, but you haven't done anything. And that's because, for digital product, the team that makes them, they carry on working on the product and carry on changing it when you as the customer, when you, the customer already has it, and so that's why the company's life, the company's interaction with the product continues even when the customer has it. So it's as if you bought a cup and then somebody from the company came to your house and they saw oh, you know what? David actually doesn't really drink that much tea, he drinks a lot more wine. So we're going to take this cup and we're going to turn it into a wine glass because that's more useful for David. That would be pretty weird if that happened, right.

08:32 - David Brown (Host)

It would be yeah.

08:34 - Sophia Matveeva (Guest)

So what happens with a digital product? And therefore digital products are never complete. And so, yes, we can all blame and be annoyed with the education system, but the education system was preparing people to thrive in one type of world. And now we have a world where there are some companies that expect perfection, because if you are buying a house, you wanna make sure that the architects and engineers have really thought this through and the ceiling is not gonna collapse. So in some parts of the economy, we expect perfection. In other parts of the economy, we expect iteration, and this is the new world where we're finding ourselves today, and this is why I think people are more confused than ever before. Because it is confusing, because you're like, okay, well, where am I supposed to be perfect and where am I supposed to bring out a minimum viable product and just see if it works? Does that make sense?

09:37 - David Brown (Host)

Yeah, totally, and it's so well that steers us. That's a good sort of on-ramp point to start thinking about AI and how AI changes that whole paradigm as well, don't you think? Because it changes the way education is gonna happen, it's gonna change the way people learn, it's gonna change the way teachers teach and I think it's gonna open up the floodgates. I mean, we're seeing this already and I use YouTube as a good example. You know, 15 years ago or 20 years ago, if you wanted to produce a video, a high quality video, that was tens of thousands of dollars worth of kit alone. You know you start thinking. You know, if you wanted to do a half an hour program once a week you're talking you know hundreds of thousands of dollars. To be able to do that, you needed professionals. You needed, you know, $50,000 cameras. You needed all this kit that you had to have.


Now anyone with an iPhone can create better quality video. Technically better quality, the style and all. That's a whole different issue. But from the camera aspect of it alone, you know what we have in our pockets now. I mean, I'm not an Apple, I use Samsung or whatever. Don't at me. You know the camera that I have in my phone right now 15 years ago was a $50,000 camera and the quality is nearly as good.


If you're a film director and you're making a feature film, no, it's not as good. But for something you know, a video for a company or something like that, like somebody's trying to film a commercial or we're doing a, you know we're doing this show. Now we're doing it remotely. I've got a 4K camera that I'm using that cost me less than $1,000. That's insane, and what that's doing is that's democratizing all of this. So this is a point I'm getting to is that not only do we have this world where you know you can do a lot of stuff digitally and you can, like you said, you're iterating and you're failing and you're changing stuff randomly all the time, but now anyone can do it, and I think that's really interesting and I don't know what your thoughts are on that. Like, where do you think that's gonna be? How? Where's that gonna take us, do you think?

12:08 - Sophia Matveeva (Guest)

she coined that term back in:


Again, 10 years later, and the differences are startling, and it already shows essentially this democratisation. So, first of all, there are now over 500 unicorns. I forgot exactly if it's kind of in the low 500s, but it's basically a 14 times increase on 10 years ago. So 14 times 14X Like even for the best venture capitalist, if they got a 14X return, they'll be really happy. So 14X increase, huge difference in what the companies do. So 10 years ago there were consumer facing companies. 80% were B2C. Today, almost 80% of the unicorns are B2B companies and they're selling enterprise software, and so this is why most people haven't heard of these companies, because they're basically selling software to large companies and, as the chief information officer at a huge corporate that's gonna be buying it, they're not household names, and here is the bit that I'm obsessed with.


Before. So in:


And I think that this makes sense because obviously we have frontier and breakthrough technologies, which are like what open AI is doing, but a lot of things. A frontier technology gets made and then essentially, it already exists and then you can basically harness it for your needs, for your business needs, and this is why I think there is this correlation. There is this rise of non-technical founders and there is more of a democratisation of what it means like who are the people who are successful in tech. Now there are more female co-founders, still not many. There are still more, I think, people called David.

16:16 - David Brown (Host)


16:17 - Sophia Matveeva (Guest)

Sorry. Yes, I forgot exactly what the names are, but there are basically three male names that you're more likely to be called one of these names than to be a female co-founder in a unicorn, but they now at least exist.


There are more non-technical founders. So we are seeing this move towards more opening and more democratisation and we're seeing it at the very, very upper echelons of tech innovation, because America basically drives how it drives the tech business innovation market. I mean, there's a lot of tech innovation in other parts of the world, but in terms of especially what we see in the West, it's really driven by the US. It's driven by Silicon Valley, and so the fact that Silicon Valley, which has, which is basically built by engineers, they are now saying you know what? Actually there's money to be made in investing in non-engineers who are leading tech companies. I think this is incredibly promising news for everybody, whether you want to be a founder or not.


And the reason why I'm so obsessed with it is when I started TechPon on TechEase the company, because I was a non-technical founder of a tech startup and there was so much stigma about being a non-technical person running a tech company and also there was nowhere I could educate myself, because it was either do a coding course or do a data science course or basically, you're not welcome here, and I was thinking I don't want to become a developer. I want to learn how to work with developers, and that's why I created a Tech for Non-Technical Founder course, which lots and lots of people have now taken. But even when I was creating it, and even as people were buying companies we're building their companies I was still thinking OK, I think I sense this thing in the market. I think I sense more and more non-technical people coming in, but is it just what I am seeing or is this part of a wider trend? And it's really hard to tell, right, because we surround ourselves, we often surround ourselves by, basically, people we like, so it's really hard to tell whether what you're seeing is the real representation. And so when this piece of data came out, I was like, oh, actually, what I'm instinctively seeing in the market is actually being reflected in the data.


And so, when it comes to AI, what I always say to non-technical founders or to basically people who want to lead in tech, is that it's not about learning to build the thing yourself, because if you try to build the thing yourself, you're not going to have time to do all the other things. You cannot run a company or you cannot be any. You can be a technical lead if this is what you're doing, but essentially, if you want to get investors or manage customers or run a marketing campaign, you can't also be the person who's coding the product. It's just not going to work. But you have to know something about technology. So learn the core concept. Understand how it works.


Like what's the difference between AI and generative AI? Why are some people still saying, oh, ai has existed since the 60s. What's all this fuss about? Like understanding what that debate is about and why people? Why we are in a revolution now, but also why there's hype. Like we're at a time when we're living in a time of AI revolution, but AI is also extremely overhyped. So how can those two things be true? Understand that work without having your own opinions. Also change your mind when you get different facts. But don't let it freak you out, because the fact that there are non-technical people leading Silicon Valley tech companies should encourage you that you can also succeed in the tech revolution and not be a techie.

20:24 - David Brown (Host)

Yeah, that's interesting, do you think that? Do you think that some of that as well has come from a maturing market? And when I say a maturing market, it's like the tech companies used to be tech and that was their USP is that they were tech right, and so you had all these engineers and stuff who had ideas about how to build new tech things. But what you've got now is everybody's move past the peak of the hype curve, let's call it. We've been through the trough of disillusionment. We're now firmly into climbing, but the focus and I think even in the time that I've been involved in and I've been working in startups since the mid-90s and back then it very much was about the tech it was all tech, tech, tech, tech. But now it's about solving problems, and so I think what's happened is that, as that's matured, people have kind of gone yeah, okay, tech, but it's great to have whizzy tech, but if you're not actually solving a business problem, then it's not worth doing the tech, and maybe that's why more non-technical people have had success, because they're the business people who come in, excuse me, and say we have this problem we need to fix, and that would also partly maybe explain the rise of the B2B companies as well.


Because I'll give you an example with AI that I've just came across the other day that I think is really interesting and it actually supports this as we were talking about voiceovers and Steve was saying that companies don't. They don't use AI for voice overs for anything because A it's so cheap, it's such a small part of the budget that it doesn't even matter to have a human do it. You don't get any financial gain from having AI do it. And he said but in some instances and I keep giving an ad for Sunglass Hut, so if Sunglass Hut wants to sponsor this podcast, that'd be fab, but Sunglass Hut has 14,000 stores in the US and so when they do their voiceover and their radio ads and stuff like that, they have a person that the talent read the ad. But now what they do is they voice print the talent's voice and they have the AI read out the 14,000 addresses, because nobody wants to read 14,000 addresses. The voice doesn't want to do it, even if they're being paid for it. They don't want to sit and just spend weeks reading addresses, right?


hat's a modern day example in:


Because I think personally I think we're past the hype cycle already on AI. I think we've reached the peak and we're rapidly sliding down into the trough of disillusionment, because I think people are starting to realise it doesn't do everything that we want it to do and people are starting to really now think about it and everybody's A lot of people I talk to they seem to be backing off a little bit, whereas they were super excited before, and I don't know if you agree with that or not. Sorry, this is a lot in there to unpack, but that was just as you were talking and explaining that. That's what it felt like to me is that it's kind of you know, the tech has moved on into business and now it's, and that's why we're seeing that. Do you think that's it? What do you think about that?

24:15 - Sophia Matveeva (Guest)

you brought somebody from the:

25:10 - David Brown (Host)

Yeah, it's crazy.

25:11 - Sophia Matveeva (Guest)

But then it took another 40 years for it to be basically pretty much everywhere. And so this is what I really want people to see is that there is a huge gulf and we've seen this with electricity fairly recently and recently enough for us to actually have the data but there is a huge gulf between a revolutionary technology being invented and it's changing the world, and essentially there are three stages. So stage one is something called a point solution, then it becomes an application solution and then a system solution, and I'll give you an example of what that means. So a good example is if you go into a factory like not an Amazon factory where everything is run by the cloud, a traditional factory, maybe even from 20 years ago. It is going to be a factory that is powered by electricity. So the entire factory is powered by electricity. There's electric lighting, the loos are going to have hand dryers that are powered by electricity and so on. But if we go back over like a century or just a bit less, they wouldn't have had electricity throughout the entire factory. Maybe it would have been kind of one process, maybe they would have had one thing which they charged through a generator and they would have had one machine that is charged through a generator. That costs a fortune but it's actually worth it. So it's kind of one thing and that's a point solution.


And the AI world. The example is a chatbot. So maybe you run a coffee shop and in order for people to book a table in your coffee shop, they have to interact with one of those chatbots that are super annoying. And it's an AI run chatbot that basically figures out, asks you a bunch of questions and then books you a table. That is a point solution, because at this level you can't say that you have an AI run coffee shop. It's just one thing that's making things easier.


Then you get to a slightly higher level of integration. So, for example, in a coffee shop, maybe the chatbot that's booking all the tables can speak to your fridge and then it can say okay, well, we're going to have a lot of reservations. We've got three birthday parties coming in on Sunday and they are probably going. Usually, when birthday parties come in, they want lots of cake and we don't have enough sugar. So in order for you to do that, okay, an order goes out automatically to buy some sugar. That's then an application solution. So there is some integration. But that's amazing and that would be super amazing if it actually worked, and that kind of already happens in Amazon Go stores, the checkout list stores, but still, when you go to the cafe, it's still going to be a human making the cake, it's going to be a human serving it to you. It's still not a completely robot operated environment, but then in several years time you could have or actually you already have some environments that are completely robot operated.


In my opinion, the way they're done right now is pretty horrible. So I don't know if any of your listeners have been to an Amazon Go store. I really don't like being in that. It's just I don't know. Even the lighting is horrible. Just the whole feeling is pretty terrible. It's kind of soulless.


It's so and you know what? I never thought that Tesco was a particularly kind of like you know, energising, soulful place where I get really like that's good for my soul. But somehow, you know, your local Tesco, like a Tesco local, in comparison to Amazon Go, just feels better.

29:02 - David Brown (Host)

Yeah, I know what you mean. I have been in one. I went in one just to try it out and it is weird, it's weird.

29:09 - Sophia Matveeva (Guest)

So even when that, even when it exists now, even when the technology exists now, there's something about it that's just unpleasant, that we don't. We don't want to be in there, and so the and Amazon is the most advanced company, right? And so what I want people to see is that right now, we're kind of in the middle times of adapting AI. The technology has been invented, the technology has been around for a while. Open AI has been working on this for almost 10 years, so this is not a new technology, but there's a huge gap, as you said, between a technology being invented and then businesses and people integrating it into their lives, accepting it into their lives, and then I mean the revolution. We will know the revolution has happened when it's not only is it everywhere, but that we don't think about it the way we don't think about it. Electricity, it's just a given.


It's just there, yeah if you came to somebody's house and they didn't have electricity, you would think that so there was something really wrong or they were really weird, right?

30:18 - David Brown (Host)

Yeah, yeah, yeah, yeah. No, you're right, and electricity is a great example to make the point. So, yeah, you're absolutely right. What do you think about? So what you started to mix into that conversation, which is is something else that I'm curious about is, is what you started to do already is to mix robotics with AI, and I think a lot of people do that Instinctively. First, I think, because we're so used to seeing robots using, like, live robots in movies and things like that. So you've got, you know, ai and she and all that other like, and you've got all these things where they have, like, a physical body. So people tend to link the two together, but I think they're still completely separate issues. I think a robot doing something and an AI doing something are still two completely different and independent things, do you? Is that how you see it, or do you really see it, as we're just waiting for the convergence of the two?

31:20 - Sophia Matveeva (Guest)

Well, so for me, an algorithm is a robot. So it's about definitions right. For me, a robot does not necessarily have to be a hardware thing. So an algorithm is a machine. Right, a machine is robot. But you know, maybe that's an incorrect definition, but I do take a point that there is difference between the actual hardware and software, and Hardware is really hard to do.


So I recently interviewed an industrial designer on my podcast and he literally said this phrase it's a miracle things get made at all, because you know, he literally talked about things like so he's in Silicon Valley and he said that you know they would design a prototype and you know they've got funding, like things are good.


Then they would get prototypes made, usually in China. The prototypes come back. Okay, maybe there are some problems, but even if everything is fantastic, what can happen and happens a lot is that your shipment from China to Silicon Valley can literally drown in the sea, and that's it. And you know it's one of these things that okay, no matter what amount of technology or AI or whatever it is you have, if the waves are high, they're not getting your stuff and your thing is not getting made, which I think is just kind of actually a Wonderful, humbling thing for us to remember, because I think we as humans, especially with AI, we can think, oh you, that's it, we can rule the world, but I think the world, just like with Kobe, can some day say no, no, no, I'm in charge.


So I think robotics, in terms of, you know, hardware, it's so difficult to do, and it's difficult to do for all sorts of reasons that are not actually to do with technology. So, for example, factories don't really like working with startups because they don't like doing small batch sizes. Factories love working with Apple because they know it's gonna be a big order. So if you've invented a new thing and you're a startup, even with Silicon Valley funding, it's still going to be really hard for you to get the best factories to even Look in your direction, right, and so this is why, whenever I see a founder who wants to, you know, make hardware, always think do you really, are you sure there are?


There are easier things to do with your life, equally with investors. Yeah, you can. I mean, apple clearly makes a lot of money. The people who invested in Apple did really really well, but so did the people who invested in Facebook, and there was a lot less risk. So so, yes, the two things hardware and software are definitely separate Software is simpler to do and hardware is difficult for all sorts of reasons. That and only a small number of those reasons are to do with technology. A lot of them are just the human factor in the business factor.

34:18 - David Brown (Host)

Yeah, and the, the physical bit also has that annoying problem with physics as well. So you know, has a bunch of constraints on it that software doesn't have yeah.


And it's like you said, it's harder. I think you know one of the things that's come up through the conversation as well. As you know, software is it's easy Compared to everything else, and what's happened. You know I was thinking about the democratization of of tech and that sort of thing. But also you know, when I was younger, if you wanted to start a business, you needed to have sort of 20, 30, 50 thousand dollars to put up front in cash, like you had to have cash flow to get that business up and running. You had to buy all your equipment, you had to have a place, you had to have a store. You know you probably had to have a couple of employees to start off with.


So the barrier to entry of Just starting any sort of a company was really really high, whereas you know, I mean I can go and start to, you know, build my own product now from the comfort of my own home. I mean I've got a technical co-founder and he and I are working on a thing and we're building our own social media Platform, sort of like Twitter. But we have a. You know we think we have something that's a special usp that people might like. We can do that in our spare time. Now it doesn't cost. It doesn't cost any capital for us to do that, other than maybe a subscription to one or two sort of you know tools that we can use to help us build it and a and a hosting platform where we can put the you know, we can put the website, and so that's also a major factor in the software side. And and AI is only again is only making that easier, because now you've got all the AI tools that Any person can just go away and say hey.


I'd like a Chrome extension that does this particular thing. You can ask chat GPT for it. It will write the Chrome extension for you and it will give you instructions on how to set it up and how to add it into Chrome. And it's like I've never done that myself. I'm not a software engineer. I know a little bit of sequel. I can run sequel queries and I can do like data querying and that sort of thing, but I can't write any code and I did it just as a test back in the beginning and in five minutes I had a Chrome extension that it didn't do anything majorly important, it was just more the fact of could I do it?


And yeah, the robotics side is you've still got that huge barrier to entry and I think Maybe it's gonna lag behind. But well, it's definitely gonna lag behind. But I think that's not a bad thing ultimately, because I think if, if you had AI Software companies running headlong into you know whatever is happening and you had Hardware that they could put it into in some sort of form, that could really move around on its own and could, that could really Cause all sorts of problems. And at least we've only got one half of that equation at the minute. So we'll see. I mean, I love the guys that engineered art, so I keep looking over here because I'm I wanted to get the name right, but they they built the amica Robot. I don't know if you've seen it, that the the guys in the UK and it's been on all sorts of stuff, but she does really complex facial expressions and it's really cool. I'll send you a link. I'll put a link also in the show notes for everybody. But I I saw the robot at CES a couple years ago when I was there with a company and it's pretty incredible Because I call her a she because it's kind of made to a female form, so it's less threatening, I think, which that's a whole another podcast to talk about, but anyway.


But she almost has like spatial awareness. So when she's standing in, like they were giving her the daily brief on who was gonna be there and what she needed to remember and things to highlight, and I was having a chat with the other co-founder off to the side but she kept looking over, like a human would do. She kept glancing because it we'd obviously move or something and it would catch in her sort of peripheral vision. She would look over, and that was the single creepiest thing about it. It was really, really off-putting because you then you actually realise that there's a little bit more going on there than just a machine Talking to you.


So yeah, sorry that's a little bit of a rant on that, but I'm fascinated by how that's all gonna develop and and. But at this point I do like to try and just Keep people, cuz everybody that I took to you that's all worried about it seems to think that the robots are gonna take over, and I'm like it's not the robots you have to worry about, it's people you have to worry about, because people, ultimately, will be the ones that will get up to bad in the various things and they'll just use whatever technology they can to do it.

38:59 - Sophia Matveeva (Guest)

Yeah, precisely so this is, and so I will. I say to people it's not AI that's going to take your jobs, it's people who know how to use AI or people who know how to program it. And, yes, you don't have to be the programmer, but you need to learn how to question how it's made and what data goes into it. So, for example, you know an algorithm. It has to have data fed into it. And what is data? It is basically a record of the past, and so if you're feeding in a record of the past and Then that you need to ask yourself is this a past that we want to recreate in the future? In some cases it absolutely is.


You know, sometimes you've done something and it was amazing and you're really proud of yourself. You're like I rock. I want to do that again, but I don't know if I can. That's exactly when you want to have some sort of computer that helps you replicate your success. Other times, you know, we can all look at examples of humanity when you know we did some really bad things and that's not a thing that we ever want to replicate. And this is the thing with Not understanding what, not being it, not questioning the data because, you know, sometimes you put in data Okay, who are the people who we hire for this particular job? And then, okay, do you really want to replicate what you've been doing so far? Sometimes yes, a lot of the time now, yeah, I also wonder.

40:30 - David Brown (Host)

Well, it's like anything, isn't it? You know you, we don't know what the unintended consequences are yet, and we're not. We can do our best to try and guess, but we won't know until they happen, I don't know. It could be fusion. You know there's a lot of discussion about nuclear fusion and how that's going to solve all the problems of the world and it's going to be the most amazing thing ever and blah, blah, blah, blah. But who knows, if we start doing that and then, 75 years later, either there's a, there's some sort of an accident or all of a sudden we realise that there's some massive, massive problem with it, or, it's more likely, it's created a secondary problem. Because what happens if you have unlimited energy? Then you need to do that, the sort of thought game about. Okay, so if we had Effectively free energy, what's the knock-on effect of that? Like, how is that going to affect society? How is that going to affect the world? What is that going to mean? And, you know, does that then mean that you're, we're going to have all sorts of you know, overhyped development of new tools and all sorts of stuff, and maybe we get new weapons, because there hasn't, like I don't know, who knows what direction it's going to go. We don't know. But that's the kind of thing, and you know this is the people who worry about AI. This is their concern is what are the unintended consequences?


You know, we thought social media was going to be. Social media was amazing in the beginning and it was. It opened up the world and everybody got to chat and it was really good for Communication and blah, blah, blah. But then you got the fraudsters involved and you've got all the people in the cyberbullies involved and you've got you know sort of all sorts of dodgy stuff going on and you've got you know it's, it's, it's now led to what we're seeing is a lot of psychological problems and kids and everything else. That that was never intended and Nobody ever even thought about those knock-on effects and I don't think at the time that they made it, they could have predicted that. So, yeah, I don't know. I'm still a glass half full guy about the whole thing. I mean, I am positive about the technology, I love it, I use it for all sorts of stuff. What about you? That's a good point. What are you? Do you use AI regularly? Do you use any tools and what do you use?

42:42 - Sophia Matveeva (Guest)

Yeah Well, I think you know, as, as we wrap up, I'd really want listeners to know that, yes, you want to learn some theory. I do suggest you learn. You do learn some theory. It's actually not that hard. And again, you can learn a lot on the tech from techies podcast, but not like on Davis podcast on YouTube.


Like there there's a lot that you can Even consume just quite passively to just get your head around things. But don't only consume things passively. You know you can get the free version of chat GPT, so I use the paid one. But before I got the paid one, I used the free one. So like that's a really good place to start and just just get used to it, because it will really, it will really annoy you and I think that's going to be really. I think that's gonna be a really good lesson because you know when you're reading all these headlines about, like always they are going to take our jobs, and then you write your first ever chat GPT prompt and you're gonna be like, oh my god, this moron is definitely I got definitely not gonna take my job. And then you rewrite your prompt and you know the output gets marginally better and then you found yourself like writing prompts for ages and then eventually, okay, you've trained a thing, but you, what I want people to see is that that experience and getting annoyed and training it and seeing how it works, that's a worthy exercise in itself because A you are not just a passive participant reading scaremongering news. You're getting involved for free, like literally download the free thing, use it for two hours. That means you now have skin in the game and you're now more informed and then see where it's relevant, really see where you could use it and where you couldn't yet, for example. So I create a lot of content.


As you said, I write for the Harvard Business Review and the Financial Times and there is absolutely no way that a Financial Times editor is going to accept a chat DBT written article. That's just not going to happen. But for me, as somebody who writes and like I pride myself on writing well, chat DBT is not something that is ever going to replace me and I'm not worried about it. But chat DBT is useful.


When I'm sitting there and we've all had this, you're like you've got a deadline, you've got to produce something, your brain is on holiday, it's just not. You're physically there, but nothing's happening, and then you basically just ask chat DBT something about the topic, and what I normally find is that I ask chat DBT, then chat DBT gives me some kind of answer that basically a smart 16 year old would have given. Like there is some fact in there, but like there is no nuance. It's written in a very bland way and then that makes me annoyed because I'm like, oh, you're missing critical points, this is no nuance. And I'm like, oh, this is now that gets my juices flowing and now I can create something. So I see it more as I mean I hate to say it, but like an annoying assistant.

45:52 - David Brown (Host)

An intern.

45:53 - Sophia Matveeva (Guest)

Yeah, exactly, it's an intern that you know that they're trying, they're really trying. They don't get it right because they don't have the industry experience, they don't understand nuance, but you know, if you ask them to give you some research, they'll give you way more research than you thought and you're like no, this is wrong. Yeah, so it's your, it's your work experience student basically.

46:19 - David Brown (Host)

Yeah, what do you think about using it? Because I've found. Well, two, what two things. One is I think they've dialled back a little bit. I think it used to be better and they've. I think, with the copyright and all the criticisms and everything that have come out, they've actually dialled back the functionality a little bit, because it used to.


It used to write you really good copy in a lot of instances and if you asked it to do something, it would just write it for you. Now, if you ask it to do something, it tells you how you're supposed to write it, which is a totally different thing, and you have to be really aggressive with it and you have to kind of ask it in several different ways to get what you want. So that's, that's point one. And now I can't remember the other one. Oh, what was the other one? I hate it when I do that. Never mind, it'll come to me in a minute. What was it?


Oh, summarising, that's it. I find it really useful for summarising information. So if you give it text or something to look at and say, look at this, and either summarise this for me or, you know, tell me what they're trying to say, or combine all of this together and pull what are the common themes or what am I missing? A lot of times I'll write something and then I'll say what am I missing? And it will come back and it will say well, on this particular topic you could talk about, you know X, y and Z and I'm like, oh yeah, that's a good point. I probably should add something in there about that. So I'm not getting it to write for me, but a lot of times almost critiquing and working as a sort of an editor have you tried it like that?

48:01 - Sophia Matveeva (Guest)

So it does give me summaries, but so I've never used chat GPT, I've never used the output as a finished product for anything. So again, yes, I can get it to give me a summary, but then it's not something that I would ever put out, even in summary format.

48:26 - David Brown (Host)

So no, I don't mean for you to then publish the summary, but even for your own information. Do you know what I mean? Like you could give it something and say, or you could take like a survey and you can take all of the answers from a survey and you can just put them all in and say what is the summary of all of this text? Yeah, what are the patterns? And it will pull out the themes.

48:44 - Sophia Matveeva (Guest)

Absolutely yes. Well, you can do that. And also you can do that with websites. There are all sorts of plugins and chat GPT for where you can literally put in a website address and then they'll basically tell you this is what this company does. So you don't I mean because you know lots of, lots of companies have really terrible websites. When you go on them and they're like we're trying to change the world and you think, great, how are you selling toilet paper or consulting services?

49:08 - David Brown (Host)

Exactly, exactly. That's your startup mentor coming out. It's like what's your elevator pitch in five words let's go Well. It's not just startups.

49:18 - Sophia Matveeva (Guest)

You know it's bigger companies that you know say we sell professional service and you think, okay, that does not narrow it down.

49:27 - David Brown (Host)

No, the best company I think in the world are doing that kind of thing and I'm conscious of time so we'll finish up here in a second but I think one of the best companies in the world at doing that brand sort of thing is Red Bull.

49:41 - Sophia Matveeva (Guest)


49:42 - David Brown (Host)

Bull is absolutely fantastic at that because they very rarely do. They ever talk about the product itself. They never tell you what's in Red Bull. They never tell you anything.

49:51 - Sophia Matveeva (Guest)

Thank God, probably Exactly. We don't want to know.

49:54 - David Brown (Host)

There's an amazing story about the origin of Red Bull that we won't go into now, but if anybody's out there, google it and see but how, what they, how they came up with the formula and all that's quite interesting. But yeah, all they're advertising is the lifestyle bit, right? It's all like you know, extreme sports and all this stuff and people do amazing things, and I talked to someone the other day who said that, as a photojournalist, if you, if you want to, if you work for Red Bull, that the logo cannot take up more than 22.7% of the image. So the logos have to be very small because they don't, they don't, they don't want the focus to be on the logo. They want the logo to be there, but they don't, they want it to be a subtle thing. They want you to focus on what's happening?

50:42 - Sophia Matveeva (Guest)

Oh, how interesting.

50:43 - David Brown (Host)

In the image they're extremely specific because they're German, so they're like 22.7%.

50:48 - Sophia Matveeva (Guest)

So they had worked out that that was the ideal. Yeah, no 22.8. That would be too much.

50:53 - David Brown (Host)

No, he said he's done images like that and they looked at it and he's like, okay, well, I need to crop it or something. Again, it's like he said I had some amazing images, but it was 28% and they wouldn't use it.

51:04 - Sophia Matveeva (Guest)

Wow, okay, we're doing once, working for them.

51:07 - David Brown (Host)

Yeah, but it works and that's exactly the thing. But they've nailed it, whereas, like you said, many companies don't, and it's just a bunch of you know you can go to a website and not understand what they do, so I prefer to like smack you in the face and just be really speaking plain English. But there we go. Yeah, sorry, I'm conscious of time. I'm sure you've got other things to do than sit and chat with me all day. Just one quick question before you go, do you? This is my question of the year. So my question of the year is do you think that we should be polite to AI when we like, when we interact with chat GPT? Do you think we should be nice to it?

51:43 - Sophia Matveeva (Guest)

Yes, I do, and it's not because of Chat GPT. I think it's because it's for the human brain. It is impossible, you know, for our brain to differentiate. I was speaking to a human or to an AI?


And if we basically just become rude and short in general, that's good. That's not an outcome that we want. And so, for example, I mean I've had this discussion with some of my friends who are parents and they've seen how their kids speak to Alexa, and some of my friends literally trained their kids to say please and thank you to Alexa and to be polite to Alexa, because if the kid is yelling at Alexa, what do you think they're going to do to kids in the playground and to other people?


And you know we want to think okay, we're not five year olds, but actually, if this is how we start speaking this to a machine, this is how we're going to start behaving. So be nice, be polite and also it doesn't really cost you anything, right?

52:53 - David Brown (Host)

Well, exactly In my other point. You know, my thought on it is that every interaction we have with it is training it further, and we don't want to train it to be mean either.

53:05 - Sophia Matveeva (Guest)

Well, what was that? Yes, Right.

53:07 - David Brown (Host)

So I think you know, I think it's worth bearing that in mind. Plus, I want it. I don't want it to think that I'm mean and when it takes over it'll kill me first, so it'll go. Hey, no, that guy Dave was pretty nice, so we'll keep him around.

53:20 - Sophia Matveeva (Guest)

Well you know, keep your insurance policy.

53:24 - David Brown (Host)

Exactly, Sophia. Thank you very much for your time today. Is there anything you've got that you'd like? I mean, obviously we want everybody to go and listen to all of your old podcast episodes to catch up on all of the amazing stuff that you've been talking about, and I have listened to a few and I just didn't want this to turn into a to that chat. But is there anything else other than the podcast you'd like to promote?

53:46 - Sophia Matveeva (Guest)

Is there anything people can go to?

53:48 - David Brown (Host)

learn more. Do you have a book or something?

53:51 - Sophia Matveeva (Guest)

So there's actually a free ebook that I have lovingly created. It's on the top 10 technology concepts for business leaders and you can get it at Forward slash, speak tech or just go to and it will pop up and this might. So this is specifically for people who are in situations where technologists or, you know, maybe venture capitalists are talking, or if you're in a company going through digital transformation and you're hearing words like API and back end and front end and UX and UI, and you're sitting there and you're nodding your head while you're thinking I have no idea what's going on, but, like I've got this job, so I've got to look like I know what I'm doing. So I've been in that situation as a non-technical founder first and it's really unpleasant. So I created this guide on the top 10 concepts that generally are the ones that people hear most, and then that should get you. That should basically be a good glossary for you.

54:54 - David Brown (Host)

Amazing and I know that'll be really helpful for a lot of people. So we will, and again I'll put links to all that stuff in the show notes. So if you just want to scroll down, if you're listening to this, you can scroll down and just click straight through to it. So, Sofia, thank you very much.

55:08 - Sophia Matveeva (Guest)

Thank you, David. It's been a really interesting, wide-ranging discussion. Thank you.

55:12 - David Brown (Host)

Thank you, cheers, I'll see you soon.

55:15 - Sophia Matveeva (Guest)


55:16 - David Brown (Host)





More from YouTube