AI CoE Simulator Demo: Ten-Minute Maturity Check

Transcript

So in this demonstration I’m going to show you the AI Centre of Excellence Simulator which is a practical tool that I’ve created and is actually part of a bigger suite of tools that I use when I’m speaking to charter directors and other executives that I sometimes will be mentoring or advising on their use of AI. And what I’ve done is I’ve pulled the AI Centre of Excellence Simulator out of that toolkit and made it publicly available on my website. The tool is designed to help organisations understand exactly where they are on their AI journey and what concrete steps they need to take next. So today I’m going to show you how the simulator can transform what are fairly abstract AI readiness concepts into actionable insights in just a few minutes. And whether you’re just starting to explore AI or looking to scale existing initiatives, this tool can provide some personalised guidance based on those proven enterprise patterns that I’m seeing every day in my conversations.

So in section one you can see here that we have the maturity journey and what you can see is that everybody starts in the observing stage. And observing is obviously where you become aware of AI’s emergence and you haven’t yet taken concrete steps towards adoption. You can also click the characteristics and the key indicators drop-downs and they will tell you a bit more about the various stage that you’re in. And similarly if you use the slider here and switch into the actual AI stages of adoption, you can see we get the same sort of output at each of the AI stages of adoption stages. So here we are in experimenting and it describes where the experimenting characteristics and key indicators are. Similarly we can jump up to scaling and this is what we’ve kind of achieved AI nirvana.

And you can see here the characteristics and key indicators are slightly different. So I’m going to switch us back to observing mode and the understanding your current stage section is really evaluating where you are in terms of what you need to do to get to transition. So this is really giving you what criteria we would consider to be mandatory and what criteria would be recommended to move to the next stage. Now we’re in the observing stage so let’s say for example we’ve got some strategic decisions to explore AI but actually we’ve now made the decision. I’ll click it again and you can see that my readiness is increasing. I’ve allocated some resources and I have a sponsor named but I’m not quite ready to move to the next stage. I’ve also going to acknowledge the competitive pressure and I’m also going to hypothesize some use cases and maybe look at the AI skills.

And I’m finally going to come back to resources allocated for exploration, movements of green and I’m now ready for transition. So I can go from observing to experimenting. So I’m going to hit the experimenting tab and you can now see that the mandatory criteria and the recommended criteria have updated to a lot more entries which actually are required to move through from experimenting to the next stage which is adopting and then on to optimizing, transforming and scaling. While you’re moving up the AI stages of adoption you can also see where you are in terms of five pillars. So the five pillars are the capabilities that you have to build in your business to be able to execute on transitions through the AI stages of adoption. And you can see here that we’re at the experimenting stage so we’re fairly low in terms of the scores across the five pillars.

And if I want to drill in I can click the functions of the five pillars and this now gives me a stage by stage explanation of where I am and what I should be doing for each of the functions within the five pillars. So within governance and accountability there are protocols for managing human AI collaboration frameworks. And at the experimenting stage we would expect to be testing basic human oversight in pilot projects without any formal guidelines. And similarly in misuse and harmful content prevention we would be doing basic content filtering in pilots and relying on vendor provided safeguards to ensure that we were compliant. Then a little bit further down and we have some example use cases that are applicable to each stage. So bearing in mind we’re in the experimenting stage. So what the simulator has suggested here for all industries is that there could potentially be a document intelligence pilot.

And for retail there could be a product recommendation email pilot. And for manufacturing there could be a visual quality inspection pilot. Now the list of pilots here will vary by stage. And obviously as you move further up the AI stage of adoption you’re able to execute more complex use cases and pilots. And I have a database now of over a thousand pilot types spread across 22 industry types. And again I’ve surfaced just three industries here and a number of pilots for each industry at each of the different phases of adoption. But many, many more to go at. So for example if I click into the document intelligence pilot I can see here that I get some leading indicators. So we would measure for example the accuracy from implementing this pilot. We would be able to see a 40% reduction in pilot completions or cost per document.

And here we can see that the enterprise impact model is a five year saving from a predictive indicator perspective. We also have some signals as to how strategically aligned this would be. So for example this would be a medium on the innovation. So probably a three out of five score on innovation. A low score from a customer value perspective. So maybe a one out of five. A medium on operational excellence. A three out of five. A medium on responsible transformation. And a low on revenue margin and profit. So actually this probably isn’t a great pilot to be doing at such an early phase. Whereas if I look at the visual inspection pilot we can see here that we’ve got some four mediums and a low. We typically expect to see strategic impact of AI pilots to increase further up the AI stage of adoption we move.

And the final section of the simulator is the recommended actions part of the simulator. And this really talks about quick wins you can do. And the timeline for them. And also the effort required. And potentially the value to the organization. So whether it’s good, it’s significant or it’s transformational. And they’re just represented by the emojis I’ve got here. And then what strategic initiatives you would typically create at this stage. So in the experimenting stage we’re looking at AI experimentation labs. Which are a medium effort. But they provide significant value. And then strategic AI portfolio. Managing really a portfolio over six to nine months. Again, a bit more effort. But still of significant value to the organization. If I scroll back up and switch us into the final phase of the AI stage of adoption. Which is obviously the scaling stage.

We’ve got no work to do there in terms of are we ready. Because we’re already in that stage. We’re typically already at the enterprise wide AI adoption stage. With a mature governance framework and so on. Our five pillars capabilities are much higher this time. You can see they’re very, very established. And the use cases have become a bit more, I would say, rounded. So a manufacturing intelligence network, for example. In the case of the manufacturing client. And you can see here that we’ve got some exceptional value. Very high value scores against our strategic objectives. And you can see here that we’ve got some quick wins here. As well as some strategic initiatives. All having a transformational impact on the AI adoption in the organization in the main. You can save or load your assessments. If you just want to save it down.

You just click the save button. And that will save a file down to your computer. And similarly you can load that assessment back up if you want to come back to it. So you can keep track of your assessment across the stages. And you can access the AI Center of Excellence Simulator at mariotomas.com. In the AI section of my website just here. Thanks for listening.