E47 Simon Kriss - Demystifying AI: A journey from fear to transformation

"AI is a little bit like fine wine in that everybody knows the name. Everybody thinks it looks great, but most people still really don't understand it."

Simon Kriss

This month’s theme

This month we explore the latest trends reshaping various sectors in Australia, from technological innovations driving business growth to market disruptors demanding strategic adaptation for survival. 

Key topics include leveraging cutting-edge technologies to spur growth as well as formulating agile strategies to respond effectively to industry disruptions.

In this episode you’ll hear about

• How to educate CEOs and leaders about AI and its potential impact.

• How CEOs and C-suite leaders need to communicate their position on AI to alleviate fears and build a safe environment for experimentation.

• How to use AI to explore new boundaries and create new products.

• How to identify pain points in the organisation and prioritise AI use cases 

• Where to start your AI journey today.

Key links

Simon's website: https://simonkriss.ai

Simon's LinkedIn profile: https://www.linkedin.com/in/simonkriss/ (Simon mentioned he will send a free PDF of his book to anyone who connects with him after listening to this podcast episode)

Simon's book on Amazon: "The AI-Empowered Customer Experience" by Simon Kriss https://www.amazon.com.au/Empowered-Customer-Experience-practitioners-possibilities-ebook/dp/B0C9YCPT4J

CSIRO's Data61: https://data61.csiro.au/

Australian Government AI Ethics Principles: https://www.industry.gov.au/data-and-publications/building-australias-artificial-intelligence-capability/ai-ethics-framework/ai-ethics-principles

OpenAI: https://www.openai.com/

Anthropic: https://www.anthropic.com/

Hugging Face: https://huggingface.co/

Canva: https://www.canva.com/

Beautify: https://beautifycloud.com/

Allmybraindecor: https://allmybrain.com/

Katonic: https://katonic.ai/

Brainfish: https://brainfish.ai/

About our guest

Simon Kriss, Chief AI Officer at simonkriss.ai is a sought-after keynote speaker and consultant. He is author of “The AI Empowered Customer Experience”, hosts podcasts on CX and AI, and was named in the 2024 CX Top 50 Global Influencers to follow. 

Based in Melbourne Australia, Simon is a CX and AI futurologist who presents to audiences around the world and works with company boards and C-suite executives to help them better understand where the AI opportunities lie for their businesses.

About our host

Our host, Chris Hudson is a Teacher, Experience Designer and Founder of Company Road, helping businesses and leaders find meaning, impact and positivity.

Chris considers himself incredibly fortunate to have worked with some of the world’s most ambitious and successful companies, including Google, Mercedes-Benz, Accenture (Fjord) and Dulux, to name a small few. He continues to teach with University of Melbourne in Innovation, and Academy Xi in CX, Product Management, Design Thinking and Service Design and mentors many business leaders internationally. 

For weekly updates and to hear about the latest episodes, please subscribe to The Company Road Podcast at https://companyroad.co/podcast/

Transcript

Chris Hudson: 0:06

Hi there and welcome back to the Company Road podcast. So sometimes in this show we look to established practices and practitioners within business to help you get ahead and to make a positive impact within your organizations. And sometimes we also bring in outside perspectives on the world of work to really stretch our thinking even further. And that's why we've looked at more abstract and creative points of view that you'll have seen and heard in other episodes as well. So, Then there are other chats that see us looking more into the future for what is to come. And that's so that we as intrapreneurs or designers or whatever we're doing or whatever we identify as can prepare for what's to come in some way. So in today's chat, we're going to turn our attention to the future a little bit. But also the present, because this is a very big theme, as we peer into the mind of one of Australia's leading futurists. And I'm very excited to welcome Australia's leading voice on AI in CX, AI mentor, keynote speaker. Simon Kriss massive welcome to the show. Thanks so much for coming in for a chat this morning.

Simon Kriss: 1:03

Thanks for having me on, Chris. I'm glad to be here.

Chris Hudson: 1:05

Thank you, Simon. And Simon, you're a customer experience, futurologist, thought leader, you work with company boards, C suite executives on a number of businesses and organizations out there. So you're going to have lots to say, I'm sure. And you're in the process of moving to a new role, I believe, around AI

Simon Kriss: 1:23

it's really just a new brand for what I've been doing. Trying to keep up with the trend.

Chris Hudson: 1:28

Exactly. And you also recently in 2023, I believe you, you recently also them self published the AI empowered customer experience, which is available on Amazon as a book. And that was right in the middle of the AI awakening period. As I'm calling it, and it's excellent timing to have you on the show because we're going to be talking in this month about the theme of AI and, future and emerging trends and what's next for Australian industries and so on. But before we get into that, and, trends and the signals of change and all of that stuff, maybe we start with your story and how you ended up in this role as a thought leader and also as an innovator in this space.

Simon Kriss: 2:04

Yeah, sure. I'd be happy to do that. So I've been in and around customer experience, I guess, more focused on contact centers, but around customer experience for about 35 years. I've worked for a number of major brands around the world. I lived in Hong Kong for 15 years before moving back to Australia and really worked globally. But I'd always been keeping an eye on artificial intelligence mostly just because I'm a bit of a nerd and it was a good topic to keep an eye on. But it never quite delivered. Traditional AI, Still didn't deliver much in the customer experience space, but with the advent of generative AI, and I started seeing this about two and a half years ago things started to change. Things started to become more malleable, more usable more, more use cases were being enabled. And then of course, once OpenAI kind of democratized generative AI and made it available to the world. I thought it was just the time to truly lean in and just work exclusively in this field. And so that's what I do.

Chris Hudson: 3:10

Yeah. Okay. And what's been your experience and tell us about some of the, I guess, some of the challenges of the hard parts that you've come across along your journey so far.

Simon Kriss: 3:19

Yeah, no, well, there's a lot of lessons to be learned over 35 years. But, in the AI space in particular. The biggest lesson is that AI is a little bit like fine wine in that everybody knows the name. Everybody thinks it looks great. Most people still really don't understand it. So I get a lot of executives that are saying, Oh, artificial intelligence. Yes. But they don't understand what that means. Like truly understand what that means. So a lot of my work is around educating executives. And actually that's where the book came from. Like I, I wrote that book a year ago. It was actually supposed to be just a little white paper of about five pages. And suddenly it became 30 pages and then 90 pages. And then I just decided it should be a book.​and so never did in my life that I think I would publish a book, but there I go a year ago, I published that and next month I'll be publishing my next book.

Chris Hudson: 4:24

Wow. Okay.

Simon Kriss: 4:25

So it's a, yeah, it's an interesting ride. But yeah the big lesson is just how little people still truly understand about AI and how it works and how they might leverage it.

Chris Hudson: 4:39

and in terms of leveraging it, tell us through, I guess, some of the, I mean, you talked a little bit there about the situation that you walk into when leaders they think it's a good idea, but they're wondering what they can do with it. And from there, where does the conversation usually go? And what are some of the steps that, an organization or a leader might usually take?

Simon Kriss: 4:57

The million dollar question, actually. It's funny because everybody's thoughts immediately go to exploitation. So how can we exploit this technology to. Be stronger, faster, more efficient, cheaper, all of those, lovely words before they start thinking about exploration. So how can we explore new boundaries with this? How could we create new products? How could we pivot our business? Things like that. And that's set against a backdrop of two things. One something I call FOMA. Which is a fear of moving forward because AI is big and scary and complicated and I don't know about it. So I'm just, I'm going to wait for the dust to settle. That's one of my favorite rules or lines that I hear. The other backdrop that it's set against is that Australian businesses tend to have a very low tolerance for failure. And when you're playing with a new tech and such a transformational tech, You have to make it okay to have some failure, some minor failure. And so taking the executives on that journey of understanding that there will be hiccups, they will trip over. They need to have a bit of a tolerance for that. They need to. Look at a whole pile of underpinning underlying issues that they maybe had swept under the carpet for a while around tech debt, things like that. But all of it comes back to what is your journey? And I would put this out to all of your listeners. What is the journey around AI? And for me, the journey starts with a personal journey. If you start using generative AI personally more and more, you're going to find that adapting it to your workplace becomes easier and easier. It's easier and easier to understand. So always start at home. And if you're looking for the perfect use case to start with at home I'll let you in on one of my great secrets. Anybody that has Has a child, a grandchild, a niece, a nephew, anybody like that. That's kind of, up to about age 12. One of the best things you can do is go into generative AI. Ask it to write you a bedtime story where that child is the hero of the story. Set it in whatever setting they're into right now. They might be into unicorns or knights of the round table or whatever it is. Set it in their favorite location. Include their dog's name or their favorite toy's name and make that a wise, trusted advisor. And you can even weave in things like. Include a message that talking to strangers is not advisable. Or, include a message that you should tell a grownup if you're being bullied or whatever message you want to weave in and those personalized stories. The kids absolutely love, and that's just a great starting point for anyone in their personal AI journey. Of course, the other one is tell it what's in your fridge and ask it to give you a recipe. That's always a winner too.

Chris Hudson: 8:20

Yeah, good. Good. So you're finding different uses for it. And I mean, I'm thinking about the child's story and also applications. From a business point of view. And obviously if you situate your story in the land of the stakeholders and in the land of the, the head of the product, marketing, whoever it is, whoever's doing it, you can think about it from that point of view too, because it's, there, there are different applications. If you learn it, like you say, at a personal level, you can then bring some of that storytelling obviously you'd reframe it for whatever it is you were working on, but you could think about how to bring it into the business context.

Simon Kriss: 8:52

Yeah, and that, that can happen a couple of ways, Chris. So one is you're right. You can just kind of, how else could I tell this story? How else could I get my point across or ask for funding or ask for approval? Whatever that might be. And then you can tailor it right down. So, somebody that I'm working with at the moment, they are in middle management and they wanted to put together a great presentation to ask for funding for some AI work from the CEO. And so I got them to dig into who the CEO was, what they liked, what they didn't like. Turns out the CEO is a Pink Floyd fan. And just recently was the 20th anniversary of Dark Side of the Moon. So what we had Generative AI do was take the key points from the message that we wanted to send to the CEO and relate them to the titles of the songs. On the dark side of the moon album. And that was actually the way he presented it to the CEO. And he said, the CEO sat there in the room for the whole thing, giggling the whole way through because he'd bothered to take the time to relate it. Like he still needed the content and he still needed the right thing, but it was just an interesting way to pitch his his idea internally.

Chris Hudson: 10:10

Yeah, I love it. I think. Knowing your audience and obviously adapting content and story for that audience in the business context is always going to work well. It reminds me of a conversation I had within a previous episode Sarah Kaur where, she works within CSIRO in an AI capacity as well. And, we were talking a little bit about creativity because she was she was an artist basically. And she was talking about, we're having this discussion about whether. Anyone can do this or whether only creative, more creatively thinking people can do this. And obviously, we'd like to believe that everyone has creativity in them, but to make the connection to Pink Floyd in the case, in the example that you give, it wouldn't have occurred to everybody, put it that way. So, so what are some of the things that you think would help spark ideas like that in the business

Simon Kriss: 11:00

There's two answers to that. The first answer is as you start on this personal journey, I was talking about doing things at home, Asking it to write, a quite descriptive email to the, to your child's school principal or whatever it is that you're using it for. What you're going to find is, as you do it once, and image generation is a great way to, to fast track this. Once you ask Generative AI to generate an image, you don't get what you want first time. Usually because your descriptor, your prompt, is too simple. And so you start learning how to become more descriptive and tell it more. I want the lighting to be cinematic. I want the backdrop to be dramatic. I want there to be fog on the lake or whatever it is. You start becoming more and more descriptive in your prompts. And therefore you get better responses from Generative AI. And so you definitely learn then to take that forward into the work situation. So instead of just saying, write me a two paragraph introduction to this report, you may start to get into, write me up an introduction that doesn't use superfluous words and has this type of feel to it. Or is going to generate a feeling of euphoria in the reading audience or whatever it is that you're trying to do. And I would encourage people to write a prompt and get a response. And then try rewriting it and try adjusting it again and adjusting it again. So that's part of the answer. The other half of the answer is looking at the use cases that AI might have within your business. And the simplest place to start there is start looking for pain points. And the most obvious use case for generative AI is not always the best. And the best is not always obvious. Sometimes things that sit in the long tail and what I mean by long tail is that 20 percent of the business that actually is quite clunky. So if you imagine someone like Kmart have Bunnings, have major contracts that, that they negotiate with suppliers and they have account managers and all that type of stuff. And then they have a whole pile of little suppliers. Well, all of those little suppliers are probably being under serviced. That may actually be a better generative AI use case than doing something for the large big business, especially initially. So, I think just looking around the organization and trying to find out, imagine an axis where on the left hand, a graph where the axes are on the left hand side how much business benefit this would bring. And along the bottom axis was how technically complex is this? Like, do we have the data even that we would need for this? We'll help you to start plotting what are the best use cases to go after. And then you prioritize them against risk and away you go.

Chris Hudson: 14:11

Yeah, I want to bring together a few things that you said there. So, obviously 2 out of 10. To a 10 out of 10 response. If you're a bit more accurate and a bit more savvy, but the prompting, and I think, like, like, like most people know now, the more that you work with the prompts, obviously, the more you can customize those to achieve the outcomes that you want and faster, probably. I mean, it's fast enough already. The fact that you have to iterate on it three times is still going to be faster than you having to spend weeks doing it in some

Simon Kriss: 14:37

That's right.

Chris Hudson: 14:39

Yeah, you were talking a little bit earlier in the conversation around exploitation, which I think is interesting. And in that previous point, you were also talking a little bit about how you know how the readiness in a sense could be assessed, and taking some of the steps around getting started. I wonder how it's sort of like you focused on the journey of AI or you focused on the outcome of AI and. Have you seen either of those work more powerfully than one another in getting the conversation started and getting the right people to agree it?

Simon Kriss: 15:09

I think it depends on where you are from a maturity perspective, an AI maturity perspective. If you're very early in the AI journey, which, most organizations today in Australia are, this year. And if you were there in the early stages, then it is about the journey. It's about trying to decide where are we trying to head with this? Because that will start to derive what tools you want to use. Do we want to take a journey that says we want to look at general use cases across the business and thus that might lead you more towards a, a Microsoft copilot or something of that ilk, or is your journey going to be one of picking off very specific pain point tasks within the organization that might each require a more nuanced approach? AI solution in order to do that. So it's a little bit like the difference between being a GP and, trying to assess anything in the human body versus being a specialist with ear, nose and throat or a specialist, with cancer or a specialist with whatever. So it early on, it's about the journey. Once you kind of decide and you start embarking on that journey, then it's about deliverables, it's about getting things into people's hands that actually work that they can use. And you will win the audience over very quickly that way. I want to emphasize that, I often say that generative and AI is a little bit like teenage romance. Everybody's talking about it. Everybody thinks everybody else is doing it. The reality is not many are doing it and those that are doing it are probably doing it wrong or without protection. So it's not like you've missed the boat on this. So it's okay to be just starting your generative AI journey or your AI journey at all, you'll struggle to win the audience over. Like I said, if you have things that go through four or five bits of failure very publicly, the organization will kind of step back. So you want to manage how that happens.

Chris Hudson: 17:19

Yeah. Yeah. No, I think so. I mean, I'm interested in as you're talking because, you make it, we always make it sound very quick or something you try out, in a matter of minutes or whatever. But I think that, the. the context of that query in whichever way the AI is set up. And obviously you can introduce it at different stages and for different purposes and in a project or, in, in rolling out some of those initiatives. But I think you've got to think about in a sense, what it's doing, where, and be quite deliberate about that. So, from a design point of view, you, you might be using it differently from, in the research stage, if you're doing discovery or, from an ideation point of view, if you're starting to use it to generate ideas and prototypes and other things, and that's a different application, obviously, so I'm just thinking that, you can use it, but actually having a base level understanding, if we step back and think about customer experience, for example, you need to know about the broader The broader system for how work is done and the role in which AI is probably going to play in each of those different stages, a little bit. So, so how are you seeing that working?

Simon Kriss: 18:24

I want to hit some low hanging, really simple, quick, easy fruit. And I want it to be a learning journey before we embark on bigger things. that's definitely there, right? So people personally, as you say, could use it to ideate. They could use it to, better wordsmith things that they're doing. They're very simple use cases that actually don't even require any data other than the parametric data that lives in whatever large language model you're using. The next low hanging one is where there is really often accessed, but unstructured data in your organization. This is stuff that's living in PDFs particularly,​but it might be in SharePoint or Confluence or something like that. So

Chris Hudson: 19:11

Yeah. Or

Simon Kriss: 19:11

a great candidate.

Chris Hudson: 19:12

Yeah.

Simon Kriss: 19:13

Yeah, a great candidate there would be all of your HR policies, right? Internally, internal HR policy. It's not HR data. It's not people's names and their leave balance and their home address, because that's a much higher risk. But just people trying to understand how do I do this or this? How do I apply for leave? What do I do if I. Think I need to be a whistleblower. What do I need to, whatever it is.​and so you can get those PDFs together and front end them with a generative AI interface, which makes it a whole lot easier. The other one from a CX perspective that people might resonate with is your. Call center workers may have a knowledge based system, a knowledge management system that when they ask it, it tends to do a semantic search. It looks for those words and it'll bring back, here's very much like Google. It'll bring back 10 links. Let's say here's these 10 documents. Your answer will be in there somewhere, but the agent's still got to leave trawl through those documents until they find the right thing. Well, you could probably put a generative AI front end on that and make it that people can just ask the question and get the answer. Now, you'd want to make sure you were controlling hallucination and doing all of those good things, but they're relatively easier use cases that don't require a lot of structured data or a lot of continuously moving data. They tend to be more based around unstructured data.

Chris Hudson: 20:49

Yeah, I mean, I think that seems pretty obvious. I think there's a lot in the use cases around, the simplest fixes basically around, efficiencies and where you can see things that are very manual, that feels like a great starting point for a lot of businesses and that's where a lot of work and time saving could be, could be made.

Simon Kriss: 21:06

Yep. And that's just. And that's just exploiting the tech, right? That's not really exploring it and changing the business up. That's just kind of exploiting the tech to make you a bit more efficient.

Chris Hudson: 21:17

Exactly. Yeah. I mean, it's a, it's used for deduction, really. I mean, you're looking at large data sets and understanding them better in an analytics capacity in a way. Yeah I'm also thinking a little bit about, I guess the fear because not people, a lot of people like take, even if it was really hard to run things through with 200 spreadsheets. The fact that it could be sped up. Comes across as a threat in some way, to Own skill set. I see this a lot in work that I do on a consultancy level where, if you're, if you're working with government or if you're working with, people that are still on their journey when it comes to digital transformation and there's a lot of other, systemic work that needs to be done for organizations like that to be able to kind of step up and step on. You come across people that take real pride in their work and that have been running things a certain way for a long time. And some are frustrated, would love to just drop it and move on and take the AI route. But actually others that, quite, quite fastidious almost around their working practices and how they do things within the tools that they have. And they just really stick to that. So from an adoption, but also, in terms of dropping legacy. Have you seen anything like that in the work that you've done as well?

Simon Kriss: 22:28

Yeah, I have seen dropping legacy, but, and it's a huge, but all of this really sits with the CEO. And for those, people that aren't adopting, I kind of blame the CEO. Like there was a report that came out about a week ago. That was a joint thing between Microsoft and LinkedIn and, they looked at countries all over the world. And they estimate that 75 percent of knowledge workers or office workers are using AI in some part of their job with varying degrees of frequency, but that at least half of those are not admitting to anyone that they use generative AI for fear of, someone saying, well, we don't need you then Yeah. That the CEOs are not coming out and laying down their position on AI. Right, so workers are really unsure what this means. Does this mean that if we adopt AI, I'm going to lose my job? Does it mean that I'll be retrained? Or does it mean, am I allowed to use it? Am I not allowed to use it? And so it's really incumbent upon the C suite to come out formally because, that fear only exists in a vacuum. We need the C suite leaders to come out and say, okay, here's where we are on AI. Even if that is as simple as we're evaluating it. We don't really know what the impact to our business is going to be yet. But we want you to use the tools or we want you to use the tools, but we want you to use them safely. So here's some guidelines around how to do that. Not a policy that says no, but a guideline that says, this is how you do use it safely and become efficient. Or that might be later on when they release their their responsible and ethical framework for AI, one of the tenets of it might be is that we're not going to use AI just to lower headcount. Now, any CEO that comes out and says that to their organization is going to have an organization that feels safe and secure around using AI because it's not going to be used just to lower headcount. Now, ultimately might headcount lower, Yes, it might through natural attrition, and we just don't refill those positions because we're becoming more efficient. There's a lot going on in this space. Could you imagine how wonderful it would be if in Australia, the use of AI became so prevalent, we became so efficient that all companies started telling their workers they only needed to work four days a week instead of five. Instead of downsizing and dumping people and blah blah. Imagine if they just said, you only have to work four days instead of five because we've got an efficiency gain out of AI. Now, a lot of people say that's a pipe dream, but imagine what that would do for your staff and, your attrition rate of staff, particularly in frontline positions would drop to almost zero because people would love that concept. But people aren't thinking that way. C suites are not coming out and saying anything about AI. But in those organizations where they are starting to talk, actively talk about it, we're finding adoption is much better. We're finding efficiency gains are just bubbling up from people within the organization saying, Hey, here's another use case.

Chris Hudson: 25:57

do you have a story maybe from a client that you've worked with or an organization that, that talks to maybe, some unexpected success from that point of view in terms of its adoption, either, from the point of view of the leadership, having communicated it well, or, from the organization, just adopting it well in some way.

Simon Kriss: 26:14

yeah, it's an organization though, that's not in Australia, it's in Singapore, actually.​where the CEO came from a technology background. and so he very much understood the tech, which helped.​and he it's a they deal in stationary and stationary supplies. And he came out and actively said, Look we want to use these tools. We think they're going to deliver great benefits. We want to see what those benefits are. This is not about dropping the number of people. And in actual fact, what they've wound up doing is winning more business because their sales people are freed up to do sales work and the excess capacity that they've created. They've actually put into areas of the business where they felt they weren't, they were falling down or they weren't doing the best they could do. So orders tend to get out a little quicker now, which means that their customer satisfaction ratings are going up. The phone calls get answered a little quicker, the emails get answered a little quicker. And so people are finding that business is becoming more responsive and as a result, they. Their sales have started to increase. It's still very early days for them. It's probably only six months in, but the early signals are really good.

Chris Hudson: 27:41

Yeah, interesting. I mean, that's great. Thank you for sharing that. I think that really brings it to life just in terms of how quickly progress can be made. If six months is sort of within the context, it feels like, I was talking about the AI awakening, right? Start the chat. Maybe from the last two years, you've been working out for a little longer than that, but it just feels like people coming around to it faster. So the adoption curve is sort of on the climb and, from the point of view of that, if you did set something out and you were going to track it over six months, in that six months, it could be quite possible that you as a business would still need to do something else and it feels like the rate of change is definitely getting faster. And I think part of the. To what you're saying about the CEO almost setting expectations, I think part of the barrier there has been around the way in which policies and, governance is set up from an organizational point of view. If you think back to the launch of social media or other similar things where there needs to be some guidance and it all has to be kind of ticked off and, Agreed by the lawyers and so on. It feels harder to do in the case of AI because it's like a, don't know what I'm agreed to because it's so, limitless really. So,

Simon Kriss: 28:50

yeah, And the risk adverse lawyers jump straight on that

Chris Hudson: 28:54

From a guidance point of view, are there simpler ways for CEOs to give that level of guidance without it turning into a six month policy writing exercise?

Simon Kriss: 29:02

Yeah, I think so. I think people I. I. believe in humans. I really do. And I think people want to come to work. They want to do a good job, but they do have this innate requirement to feel safe and secure. So when the CEO is going out with the message, the thing they should be going out with is talking about how do I make my people feel safe and secure in this fast paced, ever changing world of AI? Where all of the news in the media is about dumping people and people losing their jobs and blah, blah, blah. I mean, we've all seen this before, right? The internet put some people out of a job. Computers put a whole lot of people out of a job. Electricity put people out of a job. Hell, the steam engine put people out of a job. So, as these things come along this stuff is going to happen, with all these electric cars, how much longer are we going to need motor mechanics? So it, it's not that it's just happening here, but it's getting a lot of traction. And I think the CEOs need to come out and put their position on that. Even if their position is, we're still trying to understand the impact, but we want to experiment with it. And we want our people to be more effective and efficient where they can. So, let's start down that journey and we'll update you as we go.

Chris Hudson: 30:29

yeah, let's dive into that. I mean, around the CEOs and the change. You and CEOs a lot. I think there's a lot of, difference there, obviously, in how CEOs would handle this. And, there's one CEO and 999 people in the company. How does the level of influence come about that results in the change that we're describing, do you believe?

Simon Kriss: 30:49

Yeah, look, the most important thing for me is to get the leadership educated. Just get them educated on what it, what AI is. What it's not, what it can do, what it can't do what's it look like for their industry, just getting them to understand. You'd be amazed how many CEOs or C suite people have still never opened ChatGPT or Perplexity or HuggingFace or any of those tools. They've just. Turned away from them completely.​and that resonates also with company boards. So I work with some boards as an AI advisor and those board members have a very different or should have a very different set of questions that they're asking, because that's about governance rather than execution. But the C suite need to understand. What this tech is going to mean for them, for their organization, for their people. And I think right now they don't necessarily understand and they're trying to read their way through it. They're trying to read case studies that they've done. Don't really exist because nobody's been doing this long enough to have amazing case studies, but they're trying to read case studies and they're trying to read policy documents. And, the government is doing the same thing. It's trying to create policy around something that it doesn't really understand yet. AI is empirical. You can't think your way through it. You have to get your hands dirty. So the only way that these C suites are going to understand it is to start playing with it, using it themselves. And until we can get them to do that, they're simply not going to understand.

Chris Hudson: 32:43

so yeah, education for the CEOs anybody else in the C

Simon Kriss: 32:46

Absolutely. They need education just as much as frontline staff need education on this stuff.

Chris Hudson: 32:52

Yeah. Yeah. And then the other 999 people in the organization, the intrapreneurs, I mean, from that point of view, can they influence the adoption from the leadership? Do you think?

Simon Kriss: 33:02

I, I think they can influence the adoption. I was doing work with a a local council here in Melbourne, one of the local councils in one of the suburbs of Melbourne. And they had an internal, kind of like a hackathon, a come up with a good ideas thing. And one of the intrapreneurs in the middle of the organization actually paid his own. Subscription to an AI tool and used it to gather some information from across the organization and present it in an AI format. And the CEO was one of the people that was on the adjudication panel for this hackathon. And when he sat down and started using it. He's immediately kind of in that, Oh my God, what, why aren't we using this more in our business? Why that? So I think the trick for intrapreneurs, if they really want to push that agenda is prove it, show it, let people, like I said, it's empirical. They don't want to read about it or they can read about it, but it doesn't have the impact. Ask them to sit down in front of a PC for five minutes and find the HR policy on this. Using traditional methods or use it using this little tool that I whipped up in the last hour.​and straight away people will get it.

Chris Hudson: 34:29

Yeah. Yeah, no, I like it. I mean, I definitely see that working. I just did some work with a major bank here. And the security is super tight, obviously. And we were able to use, Gen AI to, to basically create some, it was ideation, but basically create some concept cards Not that we were showing, not putting the whole IP through the engine, but just if we wanted to show a particular image and it was doing something in a setting, a particular scene or whatever, you could bring that to life in some way through the picture. And the leadership loved it, to the extent that they wanted to be shown how it was done. And I think that's where the conversation can really start

Simon Kriss: 35:04

yep. Yeah, people, like I said I'll say it again. It's empirical. And so people learn it from a use case. They learn from doing. It's the simple reality of AI,

Chris Hudson: 35:25

Yeah, I mean, the tech adoption examples that you were giving there are interesting and AI is an interesting one, obviously, too. Just from the point of view of there's positive adoption and use cases that we're hoping to get from this, but also there's probably the more stigmatized side. Where, there'll be some people that obviously want to use it for good, but there'll be other people that either want to use it in the wrong way or, inadvertently or they're going to be using it to get around things that maybe shouldn't. So are you seeing signs of there being a misuse or misappropriation and what, what's happening in

Simon Kriss: 36:02

There are the nefarious people out there that are doing things OpenAI published a list of six different groups using that Microsoft, the naming convention, Typhoon Blue and blah, blah, blah. And they're clearly asking questions. Generative AI, how do I make a bomb or how do I do this? But they're very far and few between in the business world. What I'm seeing is accidental misuse. So people genuinely trying to do the right thing, but because they haven't been educated. Doing the wrong thing. So, people don't know, they may ask generative AI, a generic generative AI tool to write a letter to this customer explaining why we couldn't process a refund or whatever. And including the customer's name and address and everything else so that generative AI writes the entire letter. Not realizing that they shouldn't be publishing that information to an open source platform. So it's more, I don't I haven't seen. Any real, examples of people purposely knowingly going and doing that sort of stuff with generative AI. It's more being accidental or incidental just simply because they didn't know.

Chris Hudson: 37:26

that's good to know. I think, there's enough for people to figure out. It seems to be able to put it to positive use to begin with. And obviously, that, that can lead to much, much greater things and freeing up the time and so on. As you describe,

Simon Kriss: 37:40

Yeah. Look, I think people will be adults if you treat them like adults. So often when I'm writing generative AI usage guidelines for organizations, it starts off by actually talking about how they might use it. And I try to give them eight or 10 examples of things that they could do in their working day, where they could use generative AI. I'll try to give them example prompts that are relatively well written or probably better than a, a person that didn't know about generative AI might write. And I even give them the names of two or three free tools that they could, you want to build beautiful slides, great. There's Canva and there's Beautify and there's, there's Gamma, all these types of things. before I get into, okay, and if you are going to use it, here's the things you need to be cautious of. of thing. So you're not just telling them no, and smacking them over the hands, but you're actually helping them see how they would use it for themselves and let them go and explore.

Chris Hudson: 38:44

yeah, that's cool. I mean, I saw that you've done some work in the guideline space around good AI guidelines. And did you have any other advice for people that were looking at this adoption on that basis of setting up guidelines and getting some of those principles, right?

Simon Kriss: 38:58

Yeah, you definitely want to do guidelines. And if you don't have a set and you want to set. No look, there's, there is some free help out there.​CSIRO's data 61 has released some really good stuff around this. There are the eight ethical principles, ethics principles for AI from the Australian government. There, that's a guideline. And granted, a lot of that is. written or focused upon the companies that build AI, but it also applies to how people use AI. How do how does transparency and AI play out in a work environment, things like that. So there is some good resource out there. I think Australia is probably second to none in our where we are on ethical and responsible. We have some of the best AI minds in the world in this country. I don't subscribe to the theory that the U. S. is leading everything. I don't think that's true. I think all governments are struggling with legislation around this. So, yeah there's definitely stuff out there.

Chris Hudson: 40:07

Yeah, plenty of resources. That's really good to hear and encouraging from a local economy point of view that is the case as you're seeing it leading the way. Yeah, no, it's really good. I want to maybe take a look just to the future. I said at the start, we were talking about emerging trends and obviously we've spent a lot of conversation today thinking about AI and talking about its application within the business world. But, in terms of future adoption, a lot of chat today has been around present and how to get on board. But what are you seeing is coming up and some of the shifts soon. In the market or within organizations that you think businesses or entrepreneurs will need to adapt to.

Simon Kriss: 40:42

I the funny thing about AI is that, a lot of people say, well, it lowers the need for IQ and to a certain degree, that might be true, but it really heightens the need for EQ for that emotional quotient and LQ, the learning quotient. So if you want to pick up any skill, that's going to help you into the future, particularly an AI empowered future is the ability to continuously learn. Don't take for granted for one minute that, the best way to do things, or, it's always been done this way is just the craziest phrase you can ever use around an AI world. So for me, the future is going to go to the people who have the greatest ability to rapidly assimilate, adopt, and adapt what's happening around them because the changes in, AI tech are so fast. We just saw that OpenAI has released its 4. 0, not even 4. 0, 4. 0. And immediately Google replied with their new models. The book that you mentioned that I wrote a year ago, for instance, has nothing in there about multimodal and yet multimodal AI is going to become Bigger and stronger. So, I think that stuff is there. I think if you look further into the future sometimes, movies dictate reality and if you've ever seen the movie, Her,​I think we will all wind up with some form of a personalized AI assistant in our lives that understands us, understands what we like, and we don't like, the first thing in the morning, we like upbeat music, but late in the afternoon, we like classical music or whatever that might be. we like this aunt, but not that cousin. It will understand a whole lot about what we do and how we do it. It won't understand why, but it will understand what and how. And so it will become a huge assistant. And that will play out in the workspace too. I think each of us will probably have a work AI assistant that, you will choose the tone of voice. You will choose if it's male or female, German or Asian, you will choose all of that type of stuff. And it'll be someone that works alongside of you. So. And I don't think those days are far away. I could see that happening in the next 12 to 18 months. it's it's a exciting new world, but a rather daunting new world.

Chris Hudson: 43:28

Oh, I agree. I agree. I mean, we're talking I was, I had another AI expert on the show, Gareth Ryden, I don't know if you saw the episode, but yeah, we were talking a little bit about that. So if it's a four day week, then who's to say that, three day week might be after that. Well, you kind of follow the consequential train of thought through and you say, well, if the assistants are there, like you were describing, because the workforce, the workforce is basically freed up to do other things and what are they going to spend their Doing when there are so many people on the planet. Have you got any thoughts around what the consequences might

Simon Kriss: 43:59

If you think, about the eight, eight, eight model, right, that, Came out years ago, eight hours of work, eight hours of leisure and eight hours of rest. Well, if AI starts taking away a whole pile of that, then why can't that time go into leisure pursuits more? So, I would love to see it that I think more people were physically fit, more people were mentally fit. So, our working life causes a lot of stress and a lot of anxiety and a lot of pain. Having that time just to, if I said to you Chris, what would you do with four hours today? If I could just block four hours in your diary and you kind of stop and go, hang on. Maybe I'd go to the movies or maybe I'd just spend some time with my kids walking in the park, looking at the colors of flowers, or maybe I'd really finally get that jigsaw puzzle finished that I really wanted to do it's kind of that same dilemma. I think that people, when they reach retirement age and they formally retire, now, what am I going to do with this time? That's going to come earlier and earlier into our lives. And I think we will just start to adapt to that as it happens.

Chris Hudson: 45:12

know, from the point of view of human capital and the value that you place on it because if our time is free to do more value driving activities as we've been talking about then You can push yourself to do more. You can learn more. You can obviously bring more into work. And then that evolution will lead to other possibilities as well. But it feels a bit early to kind of predict what that world might look like if we were just

Simon Kriss: 45:35

It is kind of

Chris Hudson: 45:36

of our assistants. But you know what I mean?

Simon Kriss: 45:38

Oh, I could see, you spending time learning three new languages. Or something, and so I'm going to go for a trip in six months to Japan. So in the next six months, I'm going to learn basic Japanese. There we go. There'll be, there'll potentially be a ton of use cases. At some point I always get asked that question of, is AI going to become sentient and kill us all? And the answer is probably yes, but it's a long way off yet. We've still got a lot of good to get out of AI before we get to that much bad.

Chris Hudson: 46:13

yeah, no, I'm interested around the positivity that it could generate. I mean, if it was carrying the workload from a processing point of view, from an implementation point of view, and we were freed up to do other things, if we channeled ourselves into. a lot of time into other value driving activities and maybe into sustainability or, ways in which we can do things, do more of things in the real world, rather than on machines, like we are now, then that might lead to a different outcome as well. You never know.

Simon Kriss: 46:42

Carved out time, truly carved out time in our diary to just sit, think come up with new work ideas, new opportunities, new products, new services for our company. Most often when you ask. Be they intrapreneurs or senior execs in a, in an organization. When you ask them that question, it's always, I'd love to, but I'm just too busy. And they're too busy getting bogged down in stuff that they shouldn't be bogged down in. You getting copied on 20 emails that by the time you get to the, to read them all, it's self resolved anyway. All of that type of stuff, that's where AI is not designed to take our jobs. AI is just designed to take away all the cannon fodder to free us up to do much better things that only a human can do.

Chris Hudson: 47:32

I think I kind of thought it was an interesting term because it feels like in the short term, the possibilities and the, I guess the market for want of a better word when it comes to AI, it gets much more cluttered every day. So the explosion of platforms and startups are all jumping in, there's quite a lot more to navigate than there was a year ago, already. And I'm wondering from that point of view, how to. How to advise people best, we're talking about keeping it simple and starting with things, but the evolution of the possibility of the fact that it's kind of over niched already. How do people start and, how they navigate that?

Simon Kriss: 48:07

Yeah, look, I think there will be a shakeout in the industry. There'll be that this guy buys this guy and that guy buys that guy. right now, right, right now, nobody's sure, if you look at the investments being made in video, who. It's making a bundle out of selling chips is investing in hugging face and investing in anthropic and investing in this and this guy's investing in that. And so, it's still very cluttered marketplace. And so the way in which I encourage people to do that is find your first use case. across the organization, right? Do a little bit of research, spend a bit of time identifying some potential use cases, settle on your first one and get C level agreement that is the first use case.

Chris Hudson: 48:54

yeah.

Simon Kriss: 48:58

of all else, because as soon as you start stirring this pot, somebody over here in procurement wants something done. Somebody in the warehouse wants something done. Somebody over there in customer service wants something done. You have to exclude all of those. Get your first one done and delivered, even if that's with a point solution. That ultimately, you're probably going to throw away. valid because the learning that you will go through in bringing that product up, spinning it up and getting it going will be invaluable. Because your next drop probably won't be one AI use case, you'll probably drop three. And then the next time you'll probably drop seven because you get better and better at it. And once again, it's, is it a general model that you're trying to do something with? Or is it like, is your use case to make everybody in the organization write emails better? That's a big wide use case. Or is your use case that you want to get a summary of every phone call Automatically written in a consistent style and put into the CRM system. Very specific use case. So it depends which way you're going to go as to which tool set you then go and look for.

Chris Hudson: 50:17

Yeah, that's a good point because a lot of people see them all the same. But obviously, there, there are tool sets that you would use if you had any, if you had the information you're putting it in, that would be one maybe use case and another would be more where you're looking to the outside world of, almost like a search engine where you're looking outside for inputs to bring in, have you got any thoughts around that in terms of how to use what for?

Simon Kriss: 50:40

Yeah. Well, I mean, not, I want to start off by saying not everything needs a large language model, right? If all you are trying to do to go back to my previous example, if all you're trying to do is summarize CRM, Do you really need a model that's that big, it knows how to disarm a nuclear device? medium language models are where we're starting to see people really hitting. And so, and we're even seeing that for a number of the big vendors, especially in the CX space. where they might use a large language model up front to understand what the question is, but then they go and use the organization's data and use a small language model to generate the answer because they don't need the large language model for it. So it's, it, and it really is starting to become horses for courses. We're starting to see, companies that are coming out that are very broad in what they do. And these are not recommendations, but just names that spring to mind. Companies here in Australia, Katonic. Have an AI platform that can kind of be manipulated to do pretty much anything you want as against, say, another Australian company like Brainfish, which has positioned themselves very discreetly in that we answer questions for customers about your products and services. So. It really, like I said, it comes back to what is the use case that you're trying to do. And what do you think your journey is going to be over the next six to 12 months that would start to drive who you then went and talked to about a potential technical solution. But around AI, if your IT manager is, or your IT team is thinking that they're going to build it in house, it's madness. You just wouldn't, you just wouldn't build anymore. The cost, the time, the compute power have kind of put it out of reach.

Chris Hudson: 52:38

Yeah. Interesting. Because obviously, even two, three years ago, if you think about closed and open innovation systems and what sits behind that, it was obviously quite, quite a serious discussion with CTO as and when you would want to do anything like that. But from what you're saying it's not really worth putting the company walls around it and trying to do something yourself.

Simon Kriss: 52:57

it's not worth it. It just simply isn't worth it. I mean, there are ways to protect your data and get data security, cybersecurity, and all that sort of stuff, and still use tools that are commonly available, I'm often amused. With companies that question the security levels of their, of some of their providers, but not others, like, who questions the security levels of Microsoft and Salesforce and people like that, we just willy nilly use their products or they expect that, we're going to buy this AI product, are they SOC2 certified and are they GDPR compliant? And then if you go to your internal IT team and say, Are we SOC 2 certified? Are we GDPR compliant? Oh no, we haven't done that yet. So we're kind of expecting a much higher standard. Luckily most of the AI companies are still, are leaning into that and are reaching those levels of security and things, but needing to put a box around everything and have it living on your server, in your server farm. For AI, Is simply not going to be a reality at scale.

Chris Hudson: 54:05

So I just wanted to maybe finish with one question. Often we come onto the show and a lot of people that we've talked to are either entrepreneurs or have been, but in your case, slightly different. So we often ask them about their superpower. But I wanted to maybe frame the question slightly different from your point of view, which is, if. If you're to describe, I guess the superpower that the people would have if they were using this technology that you've been describing, what would it be in one way or another? But how could it enable people if you could sum that up?

Simon Kriss: 54:34

The superpower I think most people personally will get out of generative AI is. The flash speed, the speed with which you can do things, the speed with which here's a great use case. You get sent a very large report, or I get sent a lot of academic research. The ability to drop that into a generative AI model and say, please summarize this into three paragraphs that a 15 year old could understand. You don't get any big words in there. You get simple explanations of what this paper is about. And then you might decide to drill in or you decide, you know what? It's not worth investing my time in reading this when that's not actually going to help me in what I'm doing. This week.

Chris Hudson: 55:21

Yeah. Yeah.

Simon Kriss: 55:22

And so just those little things are going to make us faster and faster and faster in our work. So more than anything, I think that's the superpower.

Chris Hudson: 55:31

Yeah. Amazing. All right. Well, thanks so much, Simon. I really enjoyed the chat today and you're an excellent speaker. Obviously you've got your own podcast as well, and you've done your book and you've got another one along the way. So if people did want to get in touch and hear more about you or or ask you a question, where would they find you?

Simon Kriss: 55:46

The simplest place to find me is on LinkedIn. It's just Simon Kriss I'm the only one on LinkedIn with that name. I'm happy to say or they could go to the website, simonchris. ai and find me what I would like to do because this has been so much fun. Chris, is, anybody that is listening to your podcast, if they want to reach out to me on LinkedIn, I will send them a free PDF version of the book. They don't even have to go and buy

Chris Hudson: 56:13

Oh, amazing. Thank you. Thanks so much. That's a really kind offer. And we'll pop that in the show notes as well. And, yeah good reason to get in touch. I reckon. So check that out and, thanks so much. So I really appreciate you coming on and talking to us today.

Simon Kriss: 56:25

it. Enjoyed this a ton. It's been a lot of fun. Thanks