AI in Business: From Chaos to Competence with John Alexander Adam

0:28Boom! Shake the room, Fire Nation. about how we can catch up if we have not yet dove into AI. We'll talk about the main pitfalls to avoid and also what AI can do, but also what it can't do and oh, so much more. And a big take of a sponsor in today's episode goes to John and our sponsors. Are you ready for the ultimate all-in-one platform for entrepreneurs, marketers, coaches, and agencies? Build funnels, automate follow-ups, manage clients, and even white label your own software, say hello to our featured partner, High Level, and visit highlevelfire.com to start your free trial today. John, say what's up to Fire Nation and share something that you believe about becoming successful that most people disagree with.

1:17Hi, everybody. It's a pleasure to be here. Thanks for having me, John. So what do I think that's key to success that not a lot of people believe? I think people tend to think being successful, there's two schools of thoughts, and I don't particularly agree with either one of them. One is that success is based on coming up with something truly original, either in terms of product service or the approach to that. And the other school of thought is that success is based on frameworks. rigidly applying tried and tested frameworks. And I tend to think that the reality is a little bit different. Success tends to be about mixing and matching, synthesizing ideas that have already existed or building on them and putting them together in a unique way to solve a unique or to solve a new problem in a different context.

2:20Frameworks, I feel, they fall down because they're built-in, pristine, idealized environments. They apply to a certain set of conditions. And people tend to not question whether their circumstances, the factors affecting what they're going to do, which are often changing, those are the same conditions, factors, environment that the framework was built for. And they'll stick blindly to a framework which may not be relevant. So I think that success actually comes from adaptability and having that ability to think on your feet and to see how things are changing and to be able to synthesize and mix and match.

3:08So neither originality nor blind adherence to frameworks. Well, Fire Nation, we're going to be having a great conversation today all around AI in business from chaos to competence. And John, there's just a lot of buzz around AI right now, but I want to get real specific. I mean, how are real entrepreneurs using AI? How are they actually implementing it in the wild, so to speak? You can break applications down, I think, into four or five main use cases. I think the first one is the most obvious, it's cost savings. So it's companies that are reducing headcount by automating particular processes.

3:54Usually those are rules and fact-based processes that are pretty standardized and they're labor intensive. but they're relatively low complexity or the complexity at least follows set rules. Service improvement is the second category. So that could involve accelerating task completion or reducing errors on areas like customer claims, finance tasks, new clients, customer onboarding, just doing things in a better way that leaves the person experiencing it feeling better about what they've just been through. Productivity gains would be the third. So that can be all sorts of different applications like quickly fetching and working through information and data bridging between asynchronous systems.

4:45Getting the kind of information and putting it together would have once taken a lot of time trawling through documents in different systems, putting that information together. I think one interesting case I've seen was a financial institution where at the bottom of all of their marketing materials, they'd have to have a big chunk, which was a disclaimer, which was based on the regulations around that particular product. and every time the regulations changed a little bit, which was quite often, it would be a massive job for people to go into each and every one of those documents, which were hundreds, sometimes thousands, and update all those disclaimers, whereas now that can be automated.

5:30Fourth category is probably risk mitigation. So things like predicting system failures when failure conditions are known. So being able to preempt things which can damage revenue, reputation, all the things which are important to the companies. And I think the last category, and probably the one that's least prominent at this stage, though I do think that will change going forward, is revenue generation. So that's doing tasks that can't viably be performed by a human. and being able to do those things leads to new outcomes, new possibilities. So things like running simulations of thousands or millions of variables in areas like drugs discovery, trying different models in all different directions for all sorts of research and development.

6:24So I would say those are the five, let's say, main categories. Cost savings, service improvement, productivity gains, risk mitigation, and revenue generation. It's just crazy how many ways you can use AI today to improve your business, to become more functional, to become more efficient, to become more profitable. So if people are listening right now, they're like, man, I just haven't quite gotten there yet with AI for some reason. They haven't integrated it or adopted it to the level that they should, or maybe at all. How do they catch up? I've had this conversation numerous times over the last few months.

7:00And I think one thing which is very important for people to understand is that despite all the buzz, all the noise on social media and on platforms like LinkedIn, if you go into real companies, especially those which are midsize up, a lot of them are not really using AI a lot yet. or they're doing it in very controlled pilots. And the people within the companies, most of them are not hugely AI literate. So I don't think that's not a good thing because it is going to become more and more important and being able to do more with less has always been the secret to success in business. So learning how to do more with less via AI is going to be crucial to every company.

7:51but not being deep into the weeds yet, not having really established those AI use cases right now, it's not the end of the world. It will be in one, two, three years from now, but it isn't yet. So I think what companies do have to be doing now is they have to be active and they have to be looking for those opportunities that are perhaps not particularly extravagant, not the sexiest, most interesting use cases, but the ones where they can get those early business cases, where they can find something which is eating up a lot of resources. It is a repetitive process. It's rules-based. And go around the company, you know, looking at the different departments, the different functions, the different processes, and ask your people, you know, where are they spending their time?

8:50Identify those areas which are relatively uncomplicated to automate where ROI can be demonstrated quickly and effectively. And that does a number of things. First of all, it builds faith within the organization. It allows leaders to sell up and into the organization to unlock budget from C-suite, for example. It also gets people at all levels of the organization used to working with AI and getting them proactively involved in the process of identifying the areas where AI can be applied that helps them do their jobs better, more efficiently. You also kind of break down a little bit of that defensiveness and worry that might exist.

9:34So I think that's probably key. identify those, the low hanging fruit and do it in a way which is involving the people, involving the company, making it inclusive. So, so people don't get worried and defensive about the, the integration of AI into their workplace. Fire Nation, we have a lot to talk about around the main pitfalls to avoid what AI can do, but also what it can't do. And so much more when we get back from thanking our sponsors. You've got AI helping you write emails and name your dog. Shouldn't you be using AI to help you invest too? Seriously, I know investing can seem a bit overwhelming.

10:14I've been studying it since a young age and I still have questions 30 years later, but it's an incredibly important part of my overall long-term wealth strategy. Is it part of yours? Public.com is the investing platform for those who take it seriously. With Public, you can build a multi-asset portfolio of stocks, bonds, options, crypto, and more. You can also access industry-leading yields, like the 4.1 % APY you can earn on your cash with no fees or minimums. But what sets Public apart? AI isn't just a feature. It's woven into the entire experience, from portfolio insights to earnings call recaps.

10:47Public gives you smarter context at every touchpoint. Plus, you can earn up to$10 ,000 when you transfer your existing investment portfolio over to Public. Go to public.com slash fire to fund your account in five minutes or less. Paid for by public investing, full disclosures in podcast description. Okay, let's talk about a platform that's completely changing how entrepreneurs are running their business. And they are also our featured partner, High Level. If you're an entrepreneur, coach, agency, or course creator looking to grow faster and smarter, High Level is the all-in-one platform that replaces your entire tech stack.

11:23Seriously, think ClickFunnels, Calendly, MailChimp, Kajabi, and even your CRM all rolled up into one powerful white label solution. You can build funnels, send SMS and email campaigns, schedule appointments, manage pipelines, run memberships, and even launch your own branded mobile app. It's drag and drop simple, incredibly customizable, and built for serious growth. My favorite part, you can resell high level under your own brand. That means recurring revenue, loyal clients, and total control. So whether you're tired of duct taping tools together or ready to start offering high value services with ease, high level is your unfair advantage.

11:59On top of it all, their support team is second to none and has your back every step of the way. Seriously, every question that pops up, they will have your answer waiting. And if you visit my special partnership link, you'll get a free trial. So visit high level fire.com to start your free trial and see why thousands of businesses are switching. That's high level fire.com where marketing sales and automation finally come together. high level. Join the revolution. John, we're back. And if people are going to dive into AI headfirst, they're going to make mistakes. They're going to have some successes.

12:34They're going to have some failures. That's just part of the game when you learn something new. But what can we learn from you today about the main pitfalls that we can and both should avoid? I think one thing that I probably can't state strongly enough is when, if you're a small startup, it's a little bit different. If you're very lean, if you're a very tiny organization, you can be a little bit more carefree and experimental and try things. If you're a mid-sized company up to enterprise, there's a lot of mess. There's a lot of moving pieces that have to be accounted for. There's often regulation.

13:14There's company policy. So what we always advise is that the first thing that should be done, which will mitigate a lot of the risk going forward or a lot of the chances of failure, is to, before you do anything, is to set policy and governance rules around AI tools, AI integrations. Have that documented. You can use things like the EU AI Act. I'm sure in the USA you have a comparable legal framework or advisory framework. Get that all documented. And then anything that is done going forward, it's aligned with that governance from the beginning. Because a lot of the companies that we've worked with over the last several months, they've started doing AI pilot projects.

14:06There's not great observability. There's not a lot of measurability. So it turns out that tools are being used in ways that they shouldn't. people who have not taken enough care with data privacy, et cetera, et cetera. So make sure that structure is all established first, and then when you're building out systems and you're introducing tools, you can make sure that they're aligned with that governance that you have in place. I think the other thing which is hugely important is to build things in a modular way so that you can take things out, slot things in, avoid lock-in at all costs. These days, it's moving so fast at the moment that there's new tools, there's new models coming to market every few months.

14:51They unlock new possibilities. You don't know what's going to happen with individual tools. Are they going to be superseded? Individual providers, are they still going to be around? Is something going to go wrong? So whatever you build, make sure that different tools, different providers can be slotted in and out. There's no lock-in. And as things evolve, you build a framework that's flexible, that you can bring in the new things and you're not getting stuck five years behind because you build this really complicated system that has so many dependencies and you're locked into different things.

15:27So modularity, I would say, is a key point. Policy and government's framework to have that in place. And I think having an open, involved system where you involve wherever AI is going to be applied. The people who are working in that area should be involved in designing the AI that's going to help them, help elevate them as professionals and help make their jobs easier. So, John, thank you for breaking down some of those main pitfalls, but I would love to know what AI can do and also what it can't do right now, because I think it's important to understand its capabilities, but also its limitations.

16:08So can you break those down? What AI excels at is well-defined, high-volume, pattern-based tasks. And what it is not so good at is on the flip side. So when things are ambiguous, if they involve human intuition, a lot of edge cases where it's not a black and white answer, but it's about balancing considerations or interests, things which are not necessarily tangible and again to repeat it black and white rules based. So AI is great at things like research, it's great at parsing information. Some of the most effective applications are about taking unstructured data, converting it into structured data that can be searched, that can be synthesized, that insights can be taken out from.

17:02Anything which requires multi-step tasks, where conditions are changing, where things are ambiguous, that's where AI is not great. Also when it comes to complex human voice. So for example, AI is good at generic content, so writing product descriptions over hundreds and thousands of products. where you're, again, focused on facts and data and particular features, for example. If it's something which is about, let's say, a high-level strategy, which is broken down into a lot of different steps, going into complex environments, then it's not good at that kind of thing. So, John, you've been breaking down a lot of things for us.

17:55I mean, we've been talking about what it can do, what it can't do when it comes to AI, the main pitfalls, like how to actually adopt and integrate it into our life if we haven't done so yet. But what does the future look like? Like your finger is on the pulse. What are you seeing just over the edge? What are you seeing ahead in our future? As the new models are released, they tend to focus on two areas. One is improvements of efficiency. So being able to use smaller, less resource intensive models to again get high level results. The other are the larger models where the amount of tokens that can be processed at any one time is increasing quickly.

18:37And one of the things that AI struggles with is

18:43long processes that are more complex with instructions. And what's coming forward, people think that LLMs, chatbots, they're designed to interpret, to understand, to process natural language prompting and they can deal with it but they're actually much better when they're fed prompts information in kind of markup style so a little bit like how a computer programmer would use a markup language and that's it's something that's not hugely complex to learn but it's also not something anyone can just dive into so i think over the next several months and couple of years, language models will get better at delivering the same kind of higher quality output based on natural language prompting.

19:33And I think one of the big evolutions that we'll see will be less based on evolutions and improvements in the technology and more about the fact that they've been embedded in companies for a lot longer. So they've been ingesting the data from companies, their revenue data, their customer calls, process documentation, all of the, every kind of documentation or communication data that goes on in and around the company, combining that with news, industry analysis, those different pieces and being able to place those different pieces of the jigsaw together. So at the moment, you kind of need to put AI on fairly narrow tracks.

20:22You need to keep for it to be effective. As soon as you give it more choice, more variables, when it's not a black and white, as I've said before, when it's not a black and white, right or wrong, this way or that way, that's where it doesn't perform as well. And that's where it's going to get better over the next, not even the next several years, but the next several months, as you know, it's moving very quickly. It is crazy how fast it's moving. I mean, it's truly exponential and just seeing how AI is learning from itself. I mean, that's really the scary and crazy thing. And that's when it just goes to this compounding effect that will just never stop when the speed just continues to increase.

21:04So, John, from everything that we talked about here today, what would you say is the one key takeaway you would want our listeners to get from our conversation? Get active. I mean I know myself I work for a technology company but I'm not personally technical but I understood that this is important and a little over a year ago now I started diving into it and I started by learning how to how LLMs work the building blocks I bought a book, you can go on Amazon any bookstore, you'll find books that are that have been written for non-technical people that go into things like how to build an LLM, prompt engineering, all of those elements.

21:49And just building up that understanding of how different types of AI work. What are their limitations? What are their strengths? And experimenting. Everyone can experiment. And it's not just about typing questions into chat GPT, but set up your own custom GPT. You can do that pretty easily. fit in with documentation around the project that you're working on. Start thinking a little bit more, learn about prompting in the right way, learn the science behind that and just play with it. Do little projects, learn the building blocks and you'll be ahead of nine people out of 10 just by doing that. And I think that will be something that will position people very, very strongly in future.

22:38Well, John, if we want to connect with you, if we want to learn more about what you have going on, what is your call to action for our listeners today? I mean, people working for companies and they're interested in what AI can do in their companies, then they can get in touch. They can look me up on LinkedIn. I'm sure my contact details will also be included with the podcast. Look me up, go to our website. Let's just have a chat, you know, like for, let's, let's look at the processes, look at what the company's strategy is. What is it doing? We'll be able to find low hanging fruit together. And then you can, you can start making those, those baby steps towards the bigger steps coming down the road.

23:24Fire Nation, you're the average of the five people you spend the most time with. You've been hanging out with JAA and JLD today. So keep up the heat. For links to everything we talked about, visit eofire.com. Type John in the search bar and the show notes page will pop right up. And John, thank you for sharing your truth, your knowledge, your value with Fire Nation today. And for that, we salute you and we'll catch you on the flip side. Thank you very much. Hey, Fire Nation, a huge thank you to our sponsors and John for sponsoring today's episode and Fire Nation. You ready to rock your very own podcast?

23:56Check out our free podcasting course where I will teach you how to create and launch your podcast for free, freepodcastcourse.com. That's freepodcastcourse.com. I'll catch you there or on the flip side. Are you ready for the ultimate all-in-one platform for entrepreneurs, marketers, coaches, and agencies? Build funnels, automate follow-ups, manage clients, and even white label your own software. Say hello to our featured partner, High Level, and visit highlevelfire.com to start your free trial today.

24:34Thank you.