conversational AI

DRUID Talks Season 2 Ep#7: The Hottest AI and Automation Trends in 2024

Embark on a journey through tech's frontier with Doug Shannon in this riveting DRUID Talks episode. Explore AI, automation, and the future of IT!

Join us for an electrifying episode of DRUID Talks, where host Kieran Gilmurray embarks on a riveting dialogue with Doug Shannon, where we explore the cutting-edge advancements revolutionizing the tech sphere. From AI to automation, we dissect the pivotal shifts driving innovation and redefining industry standards.

Subscribe for the next episode!

Kieran Gilmurray:

Today, we are talking to Doug Shannon, who's a global industry leader in intelligent automation and generative AI. Doug is also a thought leader and member of multiple expert think tanks such as Gartner, Vogel, and Thea Institute.

They set the debate on what's hot and what's not hot in tech. And today we're going to talk about the hottest AI and automation trends in 2024.

Doug, welcome to DRUID Talks.

Doug Shannon:

Thanks, Karen. Happy to be here.

Kieran Gilmurray:

Doug, let's dive straight in because this is an important conversation where lots of people are wondering where to put their dollars or where to put their bets.

Doug, what do you believe are the biggest challenges facing CIOs today?

Doug Shannon:

Yeah, it's a good question. The biggest, you know, challenges are coming to CIOs and even the C suites, right? It is going to be along the lines of how fast technology is moving and the fact that it's moving so fast that by the time it comes down to your normal pipelines of CIOs where you're here, you take the idea. You normalize the idea, and then you look to actually engage in that idea. It's already passed up. It's already changed, especially with ChatGPT, and some of these other models that are coming out and these LLMs, these large world models we just had Jim, and I come out again, and so showing how that's changing and who's first, who's second, who's third, all that's coming out. Then it's a matter of, like, where do you go next?

And if you already went in one direction, do you stay on that path or do you move on to the next one? So the biggest issue that's impacting these CIOs is they're saying, like, well, what do I move, and do I make the right move or do I wait and see where it goes? So my answer to them, to give them some feedback, is you have to kind of take a look at it now. You have to learn as you go, and you have to kind of see where that's going to go in the future to kind of drive that.

So, I think it's ideal for most CIOs and C suites to build impactful changes now in your teams so that you can actually make faster movements later.

Kieran Gilmurray:

So if we're looking at this, then, Doug, because I'm sure IT leaders and CIOs are looking at the most important trends.

What are the IT trends that CIOs should be aware of or should be tracking?

Doug Shannon:

So the trends that they're seeing are whether or not to bring in AI or GenAI into the enterprise. Do they need to build their own models? And do they, or if they're not building their own models, how are they integrating other models into their current systems? And so, but if you want to take your current data, then usually this, this conversation becomes much higher than that. It's usually, you know, what kind of company are you working with? Are you a cloud-first company? Are you a mostly on-prem company? Are you a large enterprise? Are you medium or small? Depending on your company is going to depend on what you need or how you're going to integrate into this technology.

So smaller companies and medium sized companies are actually going to be more enabled to utilize GenAI than your enterprise companies. And so I think enterprise companies coming up need to realize that there's going to be a few walls they're going to hit really soon once they start diving into GenAI because they're going to find that again, they're going to come back to that question that we just mentioned is, do I do, do you build or buy? And so when you build, and you say, hey, I want that large language model, I want all my data as part of that, I want to build it on my stuff. Great. There's a high price tag there.

And then, if you want to use somebody else's model and then just ingest your information and utilize it using smaller use cases, you can do that as well. And that's a better way to learn and that's a better way to drive more impactful change inside the organization. So everybody kind of gets a feel for it, so you're ready for it. So maybe going that way first is better. If you do plan to build your own model, you actually know what you're up against.

Kieran Gilmurray:

Doug, you and I have been in what I describe as intelligent automation in the RPA space for a number of years. Is that still on-trend technology?

Do we still need RPA and intelligent automation now that we have generative AI?

Doug Shannon:

That's a big question as of late. That's the general trend with any technology. When you have new technology that comes out or new innovation that comes out, the first thing that comes into play is whether we are replacing the previous technology and whether we are enabling the previous technology to do something different. So RPA and intelligent automation, RPA in general, is probably more on the leaning side of going away because it was very simple, logic-based stuff like ins and outs, right? Intelligent automation has brought about audit controls, has brought about, you know, low code, and that's kind of what we're seeing too with GenAI, but low code in a way that these audit controls are able to make our core systems more reliable and gives us better ins and outs. And even though most people say, hey, an API is going to solve everything, it's when you actually dive into APIs, you see that they don't always solve everything, and because they don't always give you all the information that they say they do.

So some of these companies say, hey, this API gives you this, this, and that. You actually start using it, and you find out, no, I'm missing about like 30% of what is being touted here. So in that, and then you also have API leakage that happens and so you don't, with using intelligent automation, you can actually control your APIs much easier. If there are no APIs or you don't have good APIs, you can actually build out better solutions using your intelligent automation, and you actually have better audit controls over your systems. So, you know what's going in and out of those systems. I think the future really is utilizing intelligence automation and GenAI together so that your core systems are more protected and more secure while actually driving a lot of your GenAI items and use cases and things that you need to actually be more impactful as a business.

Some of the things that I go into there is the simple way to do this is really using something like, I call this, AI spanning. AI spanning is where essentially you're using your intelligent automation currently. Then you're actually enabling some of your automations to actually be involved or use GenAI to actually solve the purpose. So you're bringing in GenAI at the beginning, middle or end, so you can see drive better use cases. So, you know, also an essential where people used to talk about the people just talking about hyper-automation. Right? And hyper automation was always kind of like a weird wonky term that was kind of used almost like snake oil salesman in a way.

I usually called it hyperscaling because it made a little bit more sense. It wasn't always available or something that you hit and then came back down to refix. So, I think GenAI is going to be able to help us with AI spanning and really drive better use cases around things that weren't able to be turned into intelligent automation previously.

Kieran Gilmurray:

Yeah, and Doug, when we're talking about intelligent automation, we're talking about data movement in many regards, data interpretation. We use IDP, but largely data movement. Now, we're hearing a lot about large action models.

Would you mind giving us your definition of what large action models are and whether they will cause RPA, IA, and APIs to disappear?

Doug Shannon:

Yeah. So, large action models recently have come out and into play, especially through smaller devices, I do believe. I mean, the future of where some of that is going is because these smaller devices are being enabled with these very RPA-centric items or models that, you know, record, you know, after recording how users use something or after having thousands of recordings of understanding how something works and then building those into your large action models. Those are, those are key. And so I think that overall would be kind of like the so on, so-called, like the death of RPA.

But RPA is much different than just RPA. There's intelligent automation, and there are other things that are there. So again, like when you have large action models coming in, I think they also enable your intelligent automation teams. They also enable your intelligent automation platforms to be able to use more of your tools in a better, impactful way that really just makes it easier for the users because if the user knows what they want, it makes it an easier conversation. So, like conversation AI, for example, when you have chatbots or you have these AI agents or autonomous agents that are going to be coming out. You can actually have that conversation with this agent and say, hey, I want to do this, this, or that, or I'm ideating on an idea, and I need some help to understand, like what's the best vacation to take and how would I get there? It can actually do a lot of that because what we saw from, like, the Rabbit R1 and some of these other things that are using these large action models, we're seeing the fact that it's actually making it much easier, not just for coding, but for the end-user experience.

And so when it's easy for the end user, you're going to get adoption much faster.

Kieran Gilmurray:

I suppose that's a bit we need to remember, isn't it, that not one technology can do absolutely everything. That's the usual hype cycle. But now we have a range of technologies, you know, generative AI-infused, conversational AI, like DRUIDs. We do have RPA. I call them platforms now, much like yourself, because they do more than RPA, they are more like intelligent automation platforms and generative AI as well.

With all the technology we have, where do you see the future as people accelerate their adoption of generative AI technologies and more?

Doug Shannon:

Yeah, I mean, as we go down the path, and just as you said, like, there's no one, you know, silver bullet to solve them all, the hope is there. I think the future is the commingling of all these different technologies. So when you have intelligent automation that serves, like, your core systems, and that's how really everyone should start, they say, how do we start this path? And that path is going to be to do intelligent automation first and understand what processes you have. Drive down that path to normalize your internal processes so you understand them better.

Then bring in GenAI to say, how can we make this faster and more impactful for our users or our employees? At which point then you're kind of involving, you know, different concepts like AI spanning, and then you're involving different concepts of maybe large action models to say, hey, some of that stuff that we see with our SAP or ERP systems, let's actually use large action models to actually serve that even better, because there's that possibility there. So really what the future turns into is what I call, you know, the autonomous enterprise, where you have all of these functions used in some way, shape, or form, be it like you have a chatbot experience or an autonomous agent experience that runs a business analyst job or runs a, you know, a project manager role. So, all these agents will have different roles. They won't be taking over these people's jobs, but they're going to be enabling those same individuals to do some of the stuff that they don't need to do.

So tracking a project or tracking the action items from a project can actually build those things out and actually help track the overall process. Understanding those processes drives the business, and those processes can be automated or they can be normalized, or even enabled by using GenAI functionality as well. So really, the future is kind of coming back to the basics of, like, understanding what the business needs, understanding the goals that they're trying to hit, and then enabling the fact and being very impactful or purposeful and how things are done in the future. And so that's where the autonomous enterprise really comes into play, and you get this commingling of different technologies.

Kieran Gilmurray:

Well, see, when I hear terms, Doug, like autonomous enterprise, generative AI, large action models, and whatever else, I could be forgiven for wondering whether I spent my money now or not.

Tech is changing so fast. Should we invest, or should we hold on for a little bit longer?

Because very often, it feels like we're building the runway as the airplane is taking off.

Doug Shannon:

Yeah, I mean, that's, that's probably the biggest question at the moment, is do I stay or do I go or do I wait or do I, you know, build? I do think there's, you know, building the plane as we are taking off, trying to get off the tarmac, putting the wings on, and putting the engine on. There is learning there. I've always been a fan of, like, the see one, do one, teach one, but I always add on, learn one at the end because I want to make sure that, you know, I'm a big believer that you don't learn until you teach.

And so really, if we have teams that are trying to learn this and trying to learn how to do this in business and be more impactful and enabling their own employees to do this, you kind of have to start looking to build, and you have to start looking at, like, normalizing it, and then how does that work for us? How does that work for your own company's culture? And how are you enabling your own systems and people working together and de-siloing what that looks like? You know, a lot of the chatbots that, a lot of the things that are coming out with these autonomous agents are really looking to just, like, normalize work between teams because it would be more visible. And so bringing that stuff to light, I mean, there's some wins there to start now, but there's also cost.

Right? And so a lot of companies want to avoid spending money where they don't see the value. And right now, really, you have AI company, like, AI functions, GenAI aptitudes, and things that are available that really have a really high price tag, and it's not really ready in some regards yet. So I do think that starting off with a ton, like, using intelligent automation, building out that aspect, building up to GenAI, is going to give you more value than just trying to get the Wright Brothers airplane off the ground versus, like, building an actual, you know, a two, a two engine approach where if one fails, at least you have another way to go. So it's always good to have a backup plan.

Kieran Gilmurray:

I'm laughing there. I'm building the wrong way. You're building the airplane. It feels like everything's being built at the same time.

Doug Shannon:

This is true. It's interesting. This goes back to that question in the beginning, right? Like, what do CIOs need to worry about? They need to worry about all this.

So, do they need to worry about how the runway is being built? Who's on the runway? Do you have the right teams to support that if anything bad or good happens? And then, like, are we, do we have enough rivets, and do we have enough things to build these airplanes so that we can actually test these use cases? And if we do, what's the cost?

Because you have to understand the cost. And right now, no matter what you, what model you're on, no matter how you're doing it, enterprise cost is a lot more than like a startup or a medium sized company or something that is basically cloud first. They're going to have a lot less cost than a more on-prem-focused enterprise.

Kieran Gilmurray:

It's an interesting time. All the technology in the world is there, but where do you put your dollars?

Doug, where do you see the future of IT? What's your vision?

Doug Shannon:

The future of IT? It has needed to change its ways for a while. It needs a new facelift, right? It needs some new makeup and maybe a new hair color because IT  in the past and between IT and the business, right, IT has always had an issue. Like, you know, it's trying to keep the lights on, like classic IT, right? They're like, hey, we're here to keep the lights on. We're going to keep the servers running. You know, we're here to, you know, keep things, you know, going. And then business is saying, hey, I need this. I need this to make the business better. I need to drive sales, you know, all these things. And then, they never really wanted to work very well together. I definitely see the future of technology as commingling both. So we're going to have a better understanding of what the business needs. Business users will be more enabled because they will have the ability to use GenAI and some of these technologies to kind of have better words.

They can actually do research much easier and say, this is what I'm trying to get. This is what I'm trying to understand. And kind of in a mentor-mentee kind of relationship, just be able to take those words and build them down in a way that makes sense so they can get their ask better. And the IT person that is very tech heavy can actually have better words to be used. So I think communication is going to be a little bit better.

And I think some groups like COEs, Centers of Excellence, and the Centers of Enablement are going to be able to drive that interchange much easier. I do think that GenAI and the intelligent automation in one should still be recorded at some kind of COE because that gives you the ability to actually really have a team that is the voice of not only the IT side but it's the voice of the business side. And that interconnectivity is going to make it a much easier transition because you can also see one, do one, teach one, learn one, and you can also then drive some of these new technologies to be successful within your own environments.

Kieran Gilmurray:

Yeah, it is the tough part, isn't it, the communication bit, because the tech is there. It's understanding both sides. And it would be interesting to see that hybrid sort of role, business and technology, but I suspect experts in both fields, as ever, will be needed. It's interesting to see how technology is influencing things. Doug, you know, even today, when we're talking about GenAI, IA, RPA, APIs, large action models, you name it, we're talking about it at this moment in time. 

Things have changed since you and I spoke over the last couple of years, and they will continue to change. As you mentioned, it's very much about what it costs, how quickly I can get to market, and how I create a competitive advantage using all the technology available to us today.

Doug, thank you so much for coming to DRUID Talks. Audience, thank you so much for tuning in today. We look forward to seeing you on the next episode of DRUID Talks in the very near future.

Thank you, everyone. Thank you, Doug. 

Doug Shannon:

Thank you, Kieran. 


Stay tuned for the next episode of DRUID Talks. Subscribe to be notified at