Ep 29: Your Questions Answered: Insights on AI Integration

But I think there should be more focus on the power consumption use and that should be more open and honest. But I understand why it's not because the power demands are huge. The comparison of doing a Google search or ChatGPT search in terms of power consumption is significant, but the quality of the result is obviously better through GPT.

When actually there's a lot of really positive advancements in AI nhs. For example, you know, the amazing things that robots can now do, identifying cancers.

Before people can, and then passing those identifications on to humans for validation.

It's amazing. That's so much quicker. I'm sure all these backlogs of things within the NHS that we all know about can be expedited and yeah, it can be used for good as well. Hi, I'm Freya. Welcome to the Skynet Business in the AI Age, the podcast that dives deep into the world where artificial intelligence meets business. Join us as we explore the latest trends, challenges and opportunities in this ever evolving landscape. From innovative startups to established enterprises, we'll uncover the secrets behind the successful integration of AI technologies. So grab a cup of coffee, sit back and. And get ready to unlock the mysteries of business in the AI age.

So. Good morning, Meg.

Good morning.

Today's episode is a little bit different, isn't it?

It is, yeah.

We've got a bunch of questions that have come from the team, from our customers, and some of them have been supplemented by some evil ones I think you found with GPT. So I don't know what to expect. I don't know what you're going to ask me. And I've got to be careful I don't say the wrong thing.

Never. You can't.

Oh, I'm sure I can.

Okay. The first question, they start off a little bit more about your personal journey and then they go more into the business and sort of later on. So if you could go back and give your younger self one piece of advice, what would it be?

There's all sorts of totally politically incorrect ones I could say, but I think I'd stick with take the risks, follow the paths that are offered to you. If opportunities come up, just jump at them because you don't know where they'll take you.

Yeah. You don't grow in your comfort zone, do you?

It's much better stay doing the same thing forever. Or when somebody said, are you interested in that? And you go, well, yeah, absolutely. I'll try that and see where it goes.

Yeah, absolutely. I think it's definitely worth taking the risks, isn't it because even if it doesn't pay off immediately, you can always look back and build on things. Yeah, absolutely. Okay, so what is your overarching mission and how do you see it positively impacting the people around you?

So I've kind of got the perfect job at the moment. I get to play with toys as they're released. Our team play with more toys and share some of those findings. And then I'm able to just talk to customers about what's possible. So that lines up nicely with my mission, which is to help as many people as possible to implement AI within their businesses and lives. Because these tools are fantastic. They're evolving all the time. That just means I get to play more. So it's brilliant winning.

You get to be a big kid.

I've never grown up. It's Peter Pan.

It is a good process that we have though at the moment, isn't it? With the whole playing with different new toys and new softwares and things that are out there and testing it out so that we've got firsthand experience in what we liked about it, what we.

Didn'T like about it means that we can play with so many more things by having that blocked time every week where somebody's testing a new tool and then feeding back and having a list that we're working through. Okay, we need to check that out now. And I just added another one to the list after my drive in this morning. It's great because I don't actually have to test everything. We get the feedback on how some of these more technical, deeper tools work and how new models are working. I still play with some of them just for fun, but we're able to touch so much more and then share those findings. It's amazing.

Yeah. And it's better to have that first hand experience than just read a blog about it or something as well.

If you've not actually done it, how can you talk about it?

Yeah, absolutely. Who's the most misunderstood figure in your industry and why do you think they've been misjudged?

Well, the one that springs to mind is obviously Elon. I don't think he's helped matters with the salute he did yesterday at Trump's inauguration. We've just dated when this podcast is recorded, but I mean, it's pretty a silly thing to do.

Yeah.

But in terms of the impact he had in the starting of OpenAI, with Tesla currently being the biggest robotics company on the planet, SpaceX implementing Starlink and putting in Internet to anyone anywhere on the planet, and the grand aspirations to get to Mars. I think with hindsight, he's going to be viewed in two ways, as some sort of evil Bond villain stroking cats, but also somebody who's fundamentally changed the world. And it's easy to look at him and he's saying all these stupid things, but actually, in terms of what he's actually succeeded in implementing is game changing.

My opinion on him is quite split. I think no one who's. Who makes that big of an impact in the world is going to be liked by everybody. But I do question some of his decisions.

Yeah.

So there's. There's a mix, but I think that's going to be, you know, you're never going to be able to please everybody. We're all different. So, you know.

Yeah. And I think people have different approaches to sleep and work patterns. Forcing that on a workforce is challenging, especially morally, but the impact it's made is crazy.

Yeah, absolutely. Do you have any unique approaches to work life balance that might be considered radical by traditional standards?

There's an argument that I don't have work life balance. I think the most radical thing I probably do in terms of work life balance is being electrocuted every week at the Pulse gym. Just taking that hour out of my time to go and put on some stupid suit and go through a HIIT routine with a pt. And the impact it makes in that little amount of time is fantastic.

And get electrocuted at the same time.

Well, who doesn't like a little bit of electrocution? So that's not really part of a work life balance, but it is fitting in exercise into small amounts of time. And I think taking that time out to read and to learn what's required isn't really work life balance. It's part of my job. But making sure that fits in my calendar is critical.

Absolutely. Which future development are you not excited about? Despite everyone else's enthusiasm, there's none.

I'm excited about all of it.

Not one thing. Not one kind of potential release out there that you think that might not.

I'm excited about all of it. So, I mean, you could look at the. The use of robots in Ukraine is a horrific thing. The use of robots in war is terrible, but it's taking people away from the front line. The development of AGI, the development of super intelligence, you could go, don't do this. But I'm still excited by it because of the impact it can make. So, no, I don't think there's anything. I love technology, I love change. So I'm excited by all of it.

Good. That leads into the next question, actually. So what emerging technology or approach do you think holds the most promise for the changing of the world?

The increased reasoning and understanding capability of these models, just the leap from 4.0 to 0101 pro to 03. What we're seeing with an OpenAI's architecture and then rapidly being replicated by Deepseek out of China to give that capability of thinking is going to be incredible. And how these models are able to navigate different paths to identify potential solutions in the same way as MRNI vaccines were identified as being suitable. We'll see all sorts of different things being solved by these models. And that for me is awesome.

Staggering, isn't it? Even just the advances that I've seen since I've worked at Project Rising is insane. From just in the last year, how much things have come on, how much faster results are being presented to any query that you ask it, and just how fast the landscape is changing is huge.

It's accelerating. And even if all development stopped now, then just the capabilities that exist being implemented into so many different businesses would be like light bulbs being created to electricity being rolled out to every factory in the world. What can you do in those factories as a result of electricity? What could you do with the current level of AI if it was rolled out universally? It's amazing.

I did hear something about just the sheer amount of electricity it takes to run these newer models and it's quite a lot, isn't it?

Yeah. I saw one piece of analysis that said for producing a thousand images through AI.

That's it. Yep.

I think it was a two kilowatt heater for running for two hours. So significant draw of power. It might have been one for two hours, but that order of magnitude. And we're seeing the huge energy demands and cooling demands on these AI data centers, but I don't think that's going to stop them being developed. I think what that actually do is drive innovative approaches to cooling, drive the production of energy directly for these. And we've seen with plans to re enable a nuclear power station in the US just to power a Microsoft data center, we'll see huge, huge data centers being built and the power being put in, power supply being put in place to manage it.

Yeah, it's needed now, isn't it? It's got to that point.

I think some of the productivity gains haven't been realized yet, but the potential of that is realized and there's an absolute race on to be first to AGI and sgi.

Yeah. A lot of competition, isn't there, between all of the different companies that are.

Out there and countries?

Yeah. Okay, so if people remember you for one key idea or contribution, what would you hope it is?

I'm not sure I've contributed much yet. I think that we're just starting to help companies on that journey. So I think if an individual company can go, Project Rising helped us to achieve this. That's brilliant. And I'll take it if we can. 10x the number of companies, that is 100x it. Fantastic. But I don't think it's on me. It's on the team.

What's the boldest move you would make to reshape your industry's future and how do you want your name attached to it? You do a lot of speaking gigs, don't you, as well, and kind of have your influence on the industry?

I'd say up until 18 months ago, I didn't want my face on anything. I would quite happily have just hidden a computer. And that hasn't really changed that much. I don't really want to be tied to it. I just want to help companies through the change. And so public speaking is part of that. This podcast is part of that. I don't think you get naturally comfortable at being in front of people and being on camera, and I'm not driven to do it. It's just how we're getting our message across.

Absolutely. And I think we've really had a focus on marketing at the moment, haven't we, in the last few months? And with the Mastermind kind of just launched this month, there has needed to be a face to that, and it's made sense that it was you.

Yeah. But I think when we look to how to use AI, that doesn't need my face attached to it.

No, I suppose not.

And so hopefully that will enable us to have lots of people's faces tied to it, rather than just being me and some members of the team.

What's the biggest misconception people have about the business or the brand that you would like to clear up?

I don't know. What do you think?

I think you could make this about a misconception about AI. So what's the biggest misconception about AI?

It's all about Terminators.

That's what sprung to mind for me as well.

I get it time and time again. Project Rising AI, they're just trying to get Terminators to take over the planet.

Yeah. And that we as a company are all robots.

Well, There is that. It's been said, hasn't it? But you're going to cough in a minute just to show that you're partly human, aren't you?

Of course, yeah. I think on that, I think for a misconception you could have that companies always think if they're using ChatGPT, they're making the most of AI in their company.

We're doing AI already. Well done. Congratulations.

Yeah. Whereas actually you're just. It's the tip of the iceberg, isn't it?

One person working with an AI model on a prompt and response is not embedding capabilities into business processes. So they run autonomous.

Actually, it's a great place to start. Absolutely.

To play too, to test concepts. And we still do it, but then we look at, okay, how do we convert that exchange into something that's systemized.

And repeatable and go one step further, then another step further, then another step further and before you know it you've got a whole process that's automated instead of just one little bit. Yeah, absolutely. Which aspects of AI do you believe are fundamentally misunderstood by the general public? And how do we feel like we've already done that? What's one long term ambition for your AI services that might seem far fetched to others but completely feasible to you?

I think the toolkit. So we're in the relatively early stages of building out this toolkit which currently has three tools in it that we're testing internally and with some alpha and beta clients. But the potential of that to impact lots of businesses, to enable them to use tools incredibly easily that are going to move their businesses forward, that for me, all day long.

Yeah. It's going to be a great opportunity to get a package out there to companies when they make it easier.

How do you use these tools? What should I be doing? Or just do this, this and it's done next.

Yeah, I agree. This one's interesting. So what's one ethical challenge you faced in developing your AI services and how did you resolve? May not be our services perhaps, but ethical challenges that working with a customer we might have faced.

We've certainly been asked to work on some projects that I haven't felt comfortable with in terms of not monitoring and surveillance, but effectively elements that involve those and finding systems that were challenging conversation and ultimately we haven't taken those on. But I think also convincing people that the output of large language models can be trusted and that it's not all just hallucinations and how you put in a framework to ensure that that happens consistently when People have gone, okay, well I've used ChatGPT and it made up this load of stuff. It's a convincing bullshit machine unless you frame it and target it and take these specific actions before and afterwards. But once you've done that, the output can be trusted.

It's building trust as well, isn't it? I've spoken to a few different people now and one of the biggest questions has been how do we know it's true though? And even if it is just advice on asking GPT something, it's like, well, now the sources are there, so you can actually click on things and go and have a look for yourself.

How do you know anything's true? How do you determine what's true?

I guess you back it up. I suppose with the customers that we're speaking to, they're asking a question about their industry.

I'm asking a question about how you decide what's true or not. There's an element of trust.

Check the sources, I guess. Yeah. And test it.

Making sure that the input information is trustworthy, validated, is backed up and you can do exactly the same or ask.

It in a slightly different way, see if it comes back with a very similar thing or. I probably, I'm probably guilty of just being like, yep, okay, cool. I think for the things that I use it for in my day to day job, I've been involved in a conversation and so I was there firsthand and so if something doesn't sound quite right, it's because I was there in the conversation.

So you're doing that fact checking and so as long as you.

Oh, it's needed, isn't it back up then? Absolutely, yep. Okay, we've got a few more. What's one global problem you think AI could solve if resources and willpower were not limiting factors? World peace?

No, I'm not sure AI gives us world peace, but I think it can massively help in material sciences and that will ultimately lead to some significant wins in climate change. When you think of how we're able to make these models assess a problem, to assess potential properties of materials for a particular usage that being applied to. Okay, what's the optimum cell configuration on a solar panel? Which materials can we use to better filter salt out of water to solve irrigation in regions that are hit by climate change, I think there's just going to be huge potential there and ultimately necessity. And those two will meet.

Yeah. Something needs to change, doesn't it? But making use of the technologies that are now there and are progressing. So there's for the last, what, 10 years there's been a huge buzz around how the Earth is warming up and we need to do something about it. But in parallel to that, we've seen these huge advancements in AI all happening kind of at the same time. So why not make use of the two at the same time to kind of push things forward?

Yeah, I mean, I don't think it's a panacea. I don't think you want to give an AI the goal statement of Solve Climate Change because unless you put in some controls, the easiest way for it to solve it is to stop the people producing all of the impact in the first place. So you use AI to solve a very, very specific niche problem and then we benefit from that.

Yeah. Five years from now, what do you hope users will say about Project Rising and how Project Rising shaped their business or their lives?

When we think of our customers like partners and so I want to have been part of their success in growing their businesses significantly. And so the tools and capabilities that we're building and providing for them, just working day in, day out so they don't have to think about it and they're able to focus on building the type of business that they actually want. And we're able to go, look, we've done this thing, we did a thing with them and aren't they amazing?

We did a thing.

So I think if they look at us in five years time and are able to say Project Rising was critical to our success and we've been critical to lots of companies success and we've helped them through this transition to higher and higher levels of artificial intelligence, then move on.

I think we've already got the foundations there for that, haven't we? And it's absolutely part of all of the conversations that we have with customers and potential customers. It's all about whatever that return on investment is that somebody wants, whether it's more time in their lives for their family or whether it's a growth plan that people want to achieve. It comes up in all of the conversations that we have, doesn't it? And so we've already sort of started on that path. It's just nice to think about in 5 years time looking back at that impact, I suppose.

Yeah. I mean, you look at Neil, who we're currently building for at the moment, he's working six and a half days a week and would quite like to work five days a week. Once we've enabled that, that solution's then going to be sold to other people in the trade. And he's going to end up having a platform that enables him to retire earlier. And we're then going to have been able to help more and more people in that industry. I'm being careful not to say the industry right now, but it's, it's amazing and I think we can continue to do that. And when we're working with a company then going, okay, we've done that for you. Now how about this? We get to have fun and make people's businesses more successful.

Absolutely. And I think when we're at that early stage in a project with somebody seeing their enthusiasm, the moment they realize, okay, this is a game changer, this means I can retire early. It's a really nice thing to watch and experience somebody realize that from my perspective.

And then actually making it happen is even better.

Yeah, exactly. Definitely. What initially sparked your personal passion for AI and has that passion evolved into something else over time? I mean, it's evolved into Project Rising?

Well, I'd say it was passion for computing and so playing with computer games. Growing up on one particular game in 91 92, dune 2, which was this command and conquer style game where you were controlling different houses in obviously going too much detail for you, but in Dune environment and able to attack their installations on a map. And I discovered that by editing one file, I could create new maps in this game and completely change the layout and distribution of troops and make it more or less challenging or completely different to the original objectives. And I just got hooked on the fact that is this coding, am I a coder now or am I just hacking about with a computer to make something work? And so I went from there to study computer science and started playing with neural nets at university and thought wow, these are amazing and crap. But have then watched from the sidelines in a variety of different roles as the capabilities as really the data and the compute has increased and then and.

Hopefully the security to not be able to.

The security is obviously always improving. But then 2017 with attention is all you need. Everything changed. And so that was effectively the perfect time to start Project Rising so that we could focus on helping companies to implement AI.

Brilliant. How do you remain open to feedback while staying true to your vision for Project Rising?

I think it's really easy. I don't really know anything about anything and so I pick up little bits and see how I can use them. Within the business we take feedback from coaches, we participate in networks like bni, Mastermind and that input that we get from trusted allies, customers, networking coaches, Makes us stronger as a business and helps me be a better leader. So feedback's good. I didn't used to think so. I used to be completely the opposite way and think I know it all, but actually I know basically nothing and it's best to work with everybody else.

Do you know? That's just what I was going to say. It's better to be open to feedback as a leader of a company, isn't it, than to just think you know everything and to be quite rigid with your.

If you think you know everything in a period where change is constant and accelerating, how can you be successful?

Comes with the nature of the business, I suppose, doesn't it? With AI changing all the time, you've got to be an agile person in general.

Got to actually enjoy change, I think.

Yeah, yeah, definitely. If you were allowed to ask your audience one question, what would it be? There's too many.

The stupid side of me says, what's your favorite cheese? But I think what's my favorite question to customers is what's the biggest problem in your business? And if they're thinking about that and how they can break that problem down, how they can go through a process of five whys to find out what the root cause is and then go, okay, how can we take that problem within the business? Break it down into steps. Those steps can be automated. So just what is the biggest problem in your business and what are you going to do about fixing it? And if the answer is you're not going to do anything, then enjoy your business while it lasts. But if you're open to seeing how you can change that, then come to us and tell us what it is.

I think as well, what takes, what processes at the moment take the longest for you, what's the biggest time suck out of your day?

Yeah, that works too. But rather solve the biggest, hairiest challenge there is.

Okay, so what's the success story that best showcases the transformation? The transformative. I put my teeth back in. What's a success story that best showcases the transformative impact of your AI services on a client's operations?

I think the one that's my favorite at the moment is a company that run a call centre. They've got a lot of staff, they have really high turnover in those call centre staff because it's not a particularly fun job. It's very repetitive. And a solution that we've implemented over the past few weeks is now proving the concept of making outbound calls for one particular call type initially with two different accents so that the accent can be selected programmatically based on where the customer is in the country. That then extracts information from a conversation that feels seamless. So the short term win is just showing the stakeholders and the senior management of that organization what's possible with AI calling. The long term is going to be they're not going to have to worry about staff turnover in the call center. They'll retain the staff that want to stay and they'll be used in niche use cases and in answering calls. But the vast majority of outbound calls will be done by AI and that's going to give them a potential saving in their payroll that is 5 million huge numbers. But not by getting rid of people, just by not recruiting more people to do the same job. And so for them as a business, it's huge. But in terms of using this new capability to talk to a customer, to understand what the problem is, to then pass specific bits of data back to a system, it's huge.

Yeah, it's voice AI is, is one that I think, I think personally is overlooked quite a lot by businesses that we speak to. And I think it's one of those things that people aren't really thinking about because they just see GPT. GPT that might be what they're using so far. So when I can go in and yourself as well go into a conversation and maybe suggest voice ar, people's faces just light up because it's something that they really hadn't considered before.

It's definitely a tool we have to be careful with.

Yes.

Because you don't want to use it for cold outbound marketing because you're going to really annoy people. But where there's a legitimate business need, where it's part of your existing processes, then the potential is huge. Just the ability for one man band plumber or electrician to be able to answer the phone 247 and book in new meetings without having to worry about answering their phone. When on our job.

Yeah, that's.

That's game changing.

And how many missed opportunities must these plumbers and electricians miss out on because someone is in that much of a rush to book somebody in on the phone that they just move on to the next one on Google.

Or if you've got a problem in your house, you don't want to call around, you don't want to wait, you just want somebody there to sort it.

Yep.

And so the benefits in terms of rapid response are huge. And therefore revenue because you charge more for that emergency call out.

Yeah, absolutely. Okay, so what are some of the hidden hurdles organizations face when integrating AI solutions and how do we help them navigate these.

There's usually a quality of data issue, making sure that their information is in a format that we can use and supply into the AI models. There's usually change management adversity within the organization. There'll be somebody who wants to carry on doing it the way we've always done it. We don't need to change. So quality of data and change management.

Are the main things I think as well with the change management it's almost that that person, these people can be turned around as well. It's not an. And actually they're sometimes the people who are most then enthusiastic.

Well we can turn a technicalist into an evangelist very quickly. Just look what you can do now. Yeah, really, really, you can do it again. You don't need me to do it now.

They're the most shocked but I suppose that's potentially because they've been the most adverse to things. So haven't looked into anything, haven't experimented with any AI before. So then it's the shock factor and the, the. Yeah, the mouth, mouth open in all kind of scenario. Yeah. So do you think public fear around AI surpasses what the reality should warrant or is there genuine cause for concern?

That's deep. There's genuine cause for concern based on bad actors misusing technology. And there always has been. You think back to any major technological change, somebody somewhere has abused it for the wrong purpose.

Of course. Yeah, it's been happening.

So that continues to be a problem. I'm not concerned about a Skynet scenario where rogue AI decides to take over. But the misuse of AI is a concern and what shape that takes. And then how legislation comes in after the event to prevent it happening again will be key. And the rapid implementation of that legislation. We never used to wear seatbelts in cars. Lots of people crashed. All of a sudden it was a good idea to mandate seatbelts in cars because people died.

I think the media contributes to a lot of people's general fear. Now if I take like my family for example, they've not done a lot of research into AI, but they certainly have read the headlines. Watched the Terminator films. Exactly. So there is, there's the surface level fear from the public, but then actually if people do their research, I suppose there's a little bit more underneath as well.

There is, but ultimately what's the purpose of the press? It's to sell more reads and so the headlines are always going to be clickbaity. You're always going to be trying to pull people into content. And most of the news, as you listened this morning, commented that it's all fear and negative and generally not a good news story. Good news doesn't sell, bad news sells. So all of the media is inclined to promote negative sides of conversation, I.

Think, when actually there's a lot of really positive advancements in AI nhs. For example, you know, the amazing things that robots can now do or just the surgeries and mammograms. Yeah.

Identifying cancers before people can and then passing those identifications on to humans for validation. So instead of everything being looked at by consultants, they're analyzed and finding things that weren't being found before through AI computer vision analysis and then passing those results to be validated is increasing the early identification of cancer. That's huge.

It's amazing. And so much quicker, I'm sure. Which means then all these backlogs of things within the NHS that we all know about can be expedited and yeah, it can be used for good as well. Okay, so we've reached a point where I think it would be good to do some rapid fire AI questions.

It sounds dangerous.

Mix it up a little bit. So we've got. There's only five. And it doesn't mean you have to kind of whiz through the answer, but just the first thing that comes to mind I think would be good fun.

Right.

Okay. So the biggest current AI buzzword you wish would go away.

AGI.

Okay. Why?

Because there isn't a simple, clear definition that everybody agrees with. Different people are incentivized in different ways as to whether it's successful or not. If OpenAI declare they've achieved AGI, then they owe Microsoft a lot of money. So they're kind of inclined not to say it and are now focusing on super intelligence. It's just going to become one of these things that as we move closer and closer to general purpose intelligence, the definition will keep on tweaking and things that have already been done will just become software in the same way as it always has. It's annoying. Just because there isn't a shared common definition that people are going to stick to.

Yeah. One challenge, one AI challenge you're personally determined to solve within your services.

I've got no idea. I think our services evolve over time as the capabilities come out. And what we can't do this week we probably can do in four weeks or six weeks time, and if not, then in six months. No question. So it's just continually looking at our processes and systems and going, okay, now what? Now what can we do next to make us more efficient, to make us be able to achieve more, to be able to help more companies? So there isn't one. I thought they were quick fire.

I know industry you believe is the slowest to adopt AI but will gain the most once it does. This is tricky, but it's got to be quick.

All of the industries are slow to adopt. We're, we're ahead of early adopters generally. Most, most businesses aren't really throwing themselves into it. We obviously saw creative really getting involved early on because of the huge risk to them. I think as soon as we've got reasonably priced general purpose bipedal robots, then the level of change will be huge in every aspect. But at the moment, most industries are moving pretty slowly.

Yeah, I think things like legal accountants.

God, they're just so easy. When you think that it's. This is the rulebook using this rulebook.

Do these things solved next quickest to impact easiest. But in my opinion, I think might be one of the slowest to adopt. One positive AI story you think deserves more headlines.

I think all of the stories are getting a lot of headlines at the moment, possibly too many headlines, but I think there should be more focus on the power consumption use and that should be more open and honest. But I understand why it's not because the power demands are huge. The comparison of doing a Google search or ChatGPT search in terms of power consumption is significant, but the quality of the result is obviously better through GPT.

Your hope for how AI will transform humanity in a single sentence. Oh, this will be tricky.

You wanted quick fire. How can I quickfire a single sentence on that? Ubiquitous access to intelligence and robotic support to enable experience lifestyle. There you go, that's a sentence.

Nice. And that's the end of all the questions we have for today.

Well, that was quick fire indeed.

It was.

Awesome.

Well, thank you very much.

Well, thank you.

I think answering all the questions for me.

My pleasure. Until next time, Sam.