 I work in the dock in Dublin. So it's a global research and incubation hub that we opened just over two years ago. It's the one place globally, actually, where physically all of what we call our innovation architecture sits under the one roof. So that includes research ventures, sort of startups, and like Accenture Labs, which I lead up, which is our R&D center. Studios, which is sort of for rapid prototyping and Accenture Innovation Centers. So today, I'm going to talk to you a bit about how we can use AI to augment humanity. And I'm going to focus on mostly in the workplace. So first, I'm going to talk about where we are today with AI and why briefly. And then I'm going to talk to what I think are three myths that are going around today. And then finally, talk to three imperatives and three challenges that we face today to really be able to use AI to augment humanity. So where are we with AI? I mean, in Accenture, we've actually never seen in our history a technology with the speed of take-up and also the power to change everything we do as AI. We call it an alpha trend, partly because it actually influences other trends. So trends like blockchain and VR are all influenced by AI. And also because, as I just mentioned, it has the potential to change our work, our home life, our society, and everything. But AI has been around for a number of years. It's not going anywhere. Well, in a second, I guess. AI has been around for about over 60 years. So why today? And really, there's two key reasons. It's data, which has been talked about today, so the amount of data and the networking of data, and also cheap processing power. So that has allowed researchers to speed up and actually use algorithms and refine them to get accuracies going from about 60% sort of 15 years ago up into the 80s and 90s. And they actually become very useful then. And you can use them to help make decisions in a wide variety of situations. I don't know if the slides aren't changing. OK, yeah, this is cool. So this is sort of what I was just talking to there. So I'm going to jump on then to the three myths that go around a lot at the moment. The first one is the robots are coming for us. So this is in the sort of Arnold Schwarzenegger Terminator or HAL in 2001, that they're actually going to wipe us out. And I think this is then new. I mean, this has been around forever when something new and big comes. I mean, probably the cavemen, when they discovered fire, thought, oh, we shouldn't use that, because it could burn down our houses and villages. So therefore, should we consider not using it to keep Mormon and cook our food? I think there's definitely artificial-generate intelligence and things like that are definitely coming. But our view would be they're not going to be implemented in a way that's depicted in these movies. And if you ask around to other AI researchers, you'll get the same view. I mean, I think it's really important that we have things like open AI, where top researchers and companies that are applying AI in very cutting-edge scenarios are thinking about these things. But really, for the vast majority of us in what we're doing today, there's a lot more important things that we need to think about in terms of how AI can improve society in the way we work today. The second one is the machines will take our jobs. Again, this one isn't new. I could show you headlines from the 50s, 60s, and even the 1800s around how the machines are going to come and take our jobs. I mean, the type of jobs we do change over time. In 1800, over 80% of the world was employed in agriculture. And that's 2% today. And there were times when, actually, the rate of drop was quite quick. I mean, I think the difference of AI is there will be some areas where the speed of adoption means that jobs will change quite quickly. And that is why the imperatives and challenges I talk about are really important to tackle and put into place now. And the third one is that current approaches will still apply. So this is, oh, I can just take a little bit of AI and automate a couple of things in my organization, and I'll be fine. I mean, we believe in our research shows that actually applying AI in that way will actually make you less competitive in the long run. So in terms of using AI responsibly, the three imperatives and the three challenges. So the first imperative is re-imagine business. So this is linked to things aren't the same. So if we just use AI to automate our companies, the problem is you might get a very short-term lift to get a better ROI for a couple of years. But actually, applying AI in that way can make your company very brittle and very hard to react to change if you're only applying it to automate specific tasks. So it's really important that we think more in terms of re-engineering processes and even re-imagining the way we do business today. The next key thing is new approaches to work. So with re-imagining and redoing our processes, it's also important that we think about how we actually work. And really, our view is here that it isn't man versus machine or human versus machine. It's how do humans and machines work together as Cronin mentioned earlier. And we call that collaborative intelligence where you're using the traits and the skills of humans that humans are really good at and the traits and the skills that AI are really good at and using them together to be better together. So we have six different areas here where we see new jobs and skills being important in the new world of AI. So the first areas where humans help the machines. So in the area of training, I mean, traditionally, you'd think about, oh, a data scientist training a model. But it's a lot more than that. So for the example of chatbots, chatbots are going to be the face of the company for a large part of their customer base. And so it's really important, say, for a bank, their chatbot isn't the same as the chatbot that you log into when you are logging into a governmental agency or a media company. So actually getting the personality of the chatbot right and training it to have the personality that reflects your brand is going to become increasingly important. In the explain and sustain space, a good example of that is actually Mark Zuckerberg, when he was up in front of the US Congress and in Europe talking about the Cambridge Analytica scandal. He committed that they were hiring 20,000 new employees in the explain and sustain space. So basically to be able to explain how the AI is working and to sustain the AI, to keep an eye on it and make sure it continues to work properly. So really what he's saying is algorithms can't police algorithm. People have to complete police algorithms. And it won't just be Facebook or Google. As AI is taken up more broadly, these types of jobs will be important in every company that's using AI. And then the other area is where machines help humans. And this is sort of the more interesting area, I think. So three places where we have examples of this is in the embody space. So embody means where the AI is physically able to help humans by doing tasks. And an example is Mercedes-Benz in Germany had a factory floor where they put robots in working with the machines. And actually, initially, they put in 80% robots and 20% humans. But they actually found that they actually had to swap the balance. And there was 20% robots and 80% humans because they needed the humans to be able to actually adjust where the robot couldn't. The second area is interact. So this is where the machine actually helped humans interact. An example of this is actually Accenture built for our visually impaired employees, a pair of glasses where we used video analytics and computer vision so that when a person is wearing them and walks into a room, it actually tells them what the glasses are seeing in the room and what's in front of them. And actually, so if they go into a restaurant, they look down at a menu, it actually can read out for them what's on the menu. And the third area is Amplify. And this is actually where we're seeing a lot and where a lot of the stuff we're building and researching is being used. So an example of this is actually allowing humans to be more value-add, say, in the space of, say, a police force where they have an emergency call center and AI is used, so voice-to-text analytics and then natural language processing to actually take the person who's calling in words, put them on the screen and actually highlight for the officer that's talking to them key words that they think they should go into the incident report. And this means that the officer can focus on what they're going to ask, which is calming the person on the phone and getting the essential information out of them. And it also means that their incident reports are much quicker to produce and more accurate. Another area where Amplify is important is actually allowing people to do things they can today. So an example of that is in Indonesia, there is one radiologist to every quarter of a million people, whereas somewhere like the states, it's one in 10,000. So things like diagnosing cancer from scans. The medical imaging now is getting 90, 90% plus accuracy. So actually in Indonesia, that opens up jobs to people who aren't doctors to be able to diagnose people and also obviously saves people's lives. And then the third area, which is really key, is responsible or ethical AI. So this is a new sort of area that companies need to, and I think are waking up to, they have responsibilities. So I'll run through these really quickly because I'm running out of time. Accountability, so this is the fact that humans have to be accountable for any decisions made by a company. So it's not acceptable to say the AI made that decision and it's really important when AI is put in that the accountability for decisions where the AI might be involved is thought about and very clear. Transparency, so this is brought into the news with GDPR today and so about being transparent about when you make decisions with the AI of why, but it goes further than that and it's where you use different types of algorithms. So if it's a black box type algorithm, should you use it in the process you're thinking about? So actually making those sort of decisions. Honesty seems obvious, but actually AI today can be used. So for example, in autonomous cars to let it speed up past the speed limit when it knows no police are around and then slow back down when there's a police. I mean that's possible but shouldn't be done. So it's really important that AI is used honestly. Fairness, this is around bias which Cronin talked about earlier. There's an example recently where with facial recognition and in the states and the independent body had a look at the facial recognition technologies that most companies were using and it was 99% accurate on white males and it was 60% accurate on black females. So and there was no problem. The actual algorithm itself didn't have an issue. The issue was the way and the data that was used to train the algorithm. And then human centric, it's also really important as we use AI to think about how is this gonna affect my employees, my customers, society. And because it's so new, actually think about the unintended consequences as well as the intended consequences. So three challenges and I'll race through these. So I think skills and learning. So as I talked about earlier, our view is that it's not a jobs problem, it's actually a skills problem. There will be areas where AI very quickly changes the type of job people do. And it's really important that us as companies and society actually work to help those people to get the skills they need in the new world. So in Accenture, we actually believe this is the biggest challenge out there at the moment. We're actually investing a billion dollars a year in Accenture re-skilling our people. And we did a survey last year with a number of executives globally and 65% of them felt that their workforce wasn't prepared for the workforce with AI. But only 3% were investing in re-skilling their workforce. So if that continues, that's actually going to end up causing a huge societal problem. So there's a real impotence on companies to really step up there. But actually it can't only be companies, it has to be society at large. So better community learning, governments have a role. I mean, it's something we all should be working on and be cognizant of. The other, the next challenge at the moment is data. We need to do more on data. I mean, today you talk to people, most companies will say data is their biggest problem. But the volume of data that's going to be out there from IoT technologies, edge technologies is going to get huge, even bigger than today. Also, in terms of how it's used, the more AI is embedded, the more important that you understand and trust your data and are sure it's accurate because of the way it's being used in your company. And thirdly, with the regulation coming in, like GDPR, it's really important we find ways and safe ways and ways within the law to share data, to allow countries and companies to innovate using data because there is a real danger if data gets locked down and people can't share and use it, it will stifle innovation. And the final area is no finish line. So we reckon today that if all research and AI stopped, it would probably take about five years for companies to catch up with what's already been done and implemented. But it's not stopping, it's still going and it doesn't show any sign of stopping anytime soon. So actually helping our companies have a culture of innovation and continuous learning is going to be key going forward. So that's the third and final challenge I'll talk to you today. But for me, I think the future is really bright with AI. I know there's worries, but I think if people focus on those three imperatives and three challenges, it will have a successful future. And most of the concepts I talked about there are actually gone into in a lot more detail in a book published by our CTO in the summer called Human and Machine. And you can get that on Amazon. And that's me. Oh, and by the way, I'm gonna say at the beginning, we're hiring at the dock. Am I allowed to say that? Okay, thank you very much. And I'll be on the panel now. Thank you, Maeve. Have a seat, Maeve. Thank you very much.