 morning Full disclosure. I'm neither a health professional nor a data scientist. So bear with me. I think it should be entertaining nonetheless So Yes, we sit fjord Dublin studio sits in the dock is enters innovation Center It's a really interesting place as a designer to be based We are 45 designers that sit alongside large teams of data scientists Technologists but also people with deep industry expertise and by and large the projects. We're working on how touch AI or use AI in some shape or form As well as being a place for developing prototypes products and services We're also a venue where clients come in to talk to us about design strategy innovation technology They come into us to work through their innovation agendas around key strategic challenges and Over the last number of years. We've had many clients who come in Often very curious about AI Very often quite concerned about AI But most of the time they're confused about AI What it is what it does? This is a one of our designers used to ganbreeder to generate these dog flares Which I'm quite quite fond of And it's not surprising that people are confused about AI because now it's synonymous with technology Well, these people coming in are deep experts on their industry areas and we work across Everything from life sciences. We do life sciences health care, but also into areas like utilities working with public safety and those types of organizations and the challenge that they face around the confusion is It's both everywhere and nowhere at the moment. So we all know artificial intelligence is in our lives It's on our phones Categorizing people. It's navigating us around the world. It's it's by and large telling us what we should be watching or consuming from a media perspective So AI is ubiquitous But for a lot of the people who come in this isn't this isn't really seen as artificial intelligence Because the AI here is wrapped within products or services that we're using every day and I think John McCarthy already predicted this when he said as soon as it works No one calls it AI anymore and in this case, it's it makes it invisible. So that's confusing and Another thing that's confusing is it's moved it's moving beyond simple prediction and classification AI keeps evolving so a couple of years ago when Autodesk released the dreamcatcher their generative Parametric design tool people were really blown away by this and it's like wow and and since then this has extended into all sorts of different areas Like the idea of generative design for drug discovery for fashion for a whole range of different applications So that's confusing and what's also confusing is AI is now starting to have unintended consequences that people are becoming more and more aware of and I know this audience here will be very familiar With issues around algorithmic bias in relation to gender in relation to race We've also started to see real-world Incidences where over dependence on automation or over reliance on AI has led to incidents like fatalities with autonomous vehicles and It's also started to kind of touches and move into really weird places I don't know how many people have are familiar with what you're seeing here below This is one of thousands of videos that are up on on kids video platforms at the moment That essentially targets the revenue the ad revenue from kids nursery rhymes and kids nursery tropes on it and the Recommendation engines of these platforms have been gamified people have looked at and said oh well actually I can find a way To put my content up this one is just weird, but actually there are many videos up there with with sexual references and graphic references That are not suitable for children So all of this is bringing us to a place where I think Kevin Kelly sums it up quite nicely when he says We're morphing so fast that our ability to invent new things outpaces the rate with which we can civilize them And with AI this is where a lot of clients who come in looking for Technology consulting and strategy consulting these concerns are really there So how should people think about AI and I saying people here on purpose This isn't about experts and how experts think about it How should we all think about it particularly if you've got businesses that are coming in that are looking for help with their innovation We can ask an expert. What is AI I'm sitting in front of lots of experts I think if I was asked many of you I would hear it's a technical term an umbrella term for a range of machine learning Techniques and would start to talk about things like random forests and neural nets and this is all technically accurate The difficulty is for Experts and other domains when they come in they find it very difficult to relate these terms to their business challenges or to their everyday lived experiences So as designers working in an innovation center part of our responsibility is to make the technology a lot more accessible and One technique that we've used to this is essentially reframing AI in human terms So when we work on innovation workshops for clients, we don't talk about computer vision or natural language processing We use analogies we talk about how might you solve that business problem problem through seeing but seeing at scale or Through reading but reading enormous volumes or through hearing touch recommendation creation But reframing AI in human terms This isn't to dumb it down It just opens up the opportunities to allow other people to ideate around how it may be used to solve their problems And I'm going to use this framework now just to quickly step you through some work that we've been doing at the dock And we use these as inspirational examples within our ideation workshops as well. So let's look at seeing So computer vision is a hugely powerful technology By and large we've seen it applied in in various different areas. I think in in Quality control. It's got obvious applications in health. We've seen it in terms of radiology as a design agency We work a lot with consumer facing products and actually one of the projects we did was around retail We're traditionally a service design agency. That's where fjords background was So we looked at how can we improve the consumer experience in this case? Using the extra eyes that we carry in our pockets to take things off the thinking list so to make our lived experience easier I'll just play the short video here So computer vision here is just a touch point within an experience a digital experience that can improve a consumer's life We also have things like reading so natural language Processing as a technology is extremely powerful We've seen how this can be applied to things like social media But actually how we've been looking at how do we bring this back to ourselves as design as designers and researchers We spend a lot of time look at doing user research on individuals One client that came to us was an insurance client who wanted to move beyond traditional insurance products Refresh your premium once a day once a year and move into areas of digital health services The difficulty was while they had loads of data on their customers that data was all framed in terms of risk And they didn't really know what the lived experience of those people were you need that in order to build out digital health services So we did what we always did we went out and conducted research But we also use natural language processing to accelerate that process Again, just a quick video here We interviewed 200 senior care providers We analyzed their responses using natural language processing and generated by gram networks that surface the key important themes For example when asked how could your caregiver learning experience be improved? We can see that carers need support and that they want classes, but they have little time for them So they need better options for learning these visualizations provided a reference for discussion with our client around the true user needs So we discover insights, but then we bring those insights directly into the design of digital products or services So hearing digital signal processing we like we Were all a lot more familiar these days with conversational AI, but how can that actually get into work forces? We did some work with a large policing authority around their computer to aid or dispatch or around their emergency dispatch process Quite a difficult challenge for them because you've got a human who's picking up the phone listening to a very stressed other human Who's also often those people are not touch typists. They're not very deeply familiar with the technology So we looked at how can we actually use? Artificial intelligence to allow us to extract the key essential information and speed up the overall process I'll just quickly play a short example here Please emergency Hello, there was an assault in the street. There's a man down. He was attacked by two persons on a motorbike Can you tell me where this has happened? I'm in Green Street. I think he's been stabbed. Please confirm the man is in Green Street. You believe he's been stabbed Yes, yes, the man is in Green Street, and and I think they tried to rob him and he's been stabbed I'm just gonna keep moving on that but what you saw there was the use of natural language processing speech-to-text technology Certain recommendation systems not to replace a human worker here, but to augment them Touch is another interesting one So when we hear about when in public consciousness people here here AI and they think robots now AI is not robots But AI is changing the way in which robots can interact with us moving away from the purely highly dangerous activities where people were segregated from them and actually moving into the area of Collaborative robots or co-botics Working together actually with with I am or the manufacturing research group or outside the team at the dock developed this This experience that allowed people to play around with actually if I work alongside a robot Can I complete these tasks a lot more quickly? Have I more flexibility? Is it easy to train and this has been a really great example of how we can work better with robots in order to achieve achieve our goals Recommendation so Recommender engines are everywhere, you know on social on on our own media platforms It's telling us a lot of the stuff we should consume but actually with IOT and with devices out at the edge providing information back to us We can really extend what we do with recommendation systems a project that we recently finished up at the dock Ran on the title the last mile and this was dealing with the business challenges facing postal services internationally This is a huge issue. Their business model is quite Victorian It's based around sticking a letter to you know the letter box of your of your permanent abode once a day It has no relevance in a world where people want fast flexible parcel delivery So using a combination of demand prediction and route optimization We're allowing postal services to leverage their existing workforces and their infrastructure to help them to compete against Same-day delivery services and what's key here again is this isn't about replacement And it's also not dictating to everybody what they should do one of the key things about getting the workforces to accept This was to provide them with mechanisms to provide feedback to the algorithm So your delivery agents out in the field can pass back information Whether that's about problems with the delivery problems with the truck the algorithm should be doing the work around the data That isn't available to them that these human agents the human glue can give back to them and then finally create I already talked about the dreamcatcher project But the whole area of computational creativity is really exploded of the last while for everything from synthetic people True to synthetic fashion so stitchfix using AI to power fashion We've been playing around with Technology with some of our work with food producers so big food production clients that we have in we were looking at How we could help them come up with new product lines that are both Interesting and pleasant and we actually set up a little test kitchen inside and work and created a range of tapas that were inspired by by a knowledge graph The middle row there is a dark dark chocolate bool sitting on a parmesan crisp So it's kind of a chocolate taco. It was actually really really good So this this activity we call the AI creative matrix and like I said we use this Working with clients we use design thinking methods to get them to frame their problem to understand what they need to do And then we use this to get them to very rapidly come up with a range of ideas around it I'm conscious. I'm way over That is just one tool that we use within our tool kit of a program called designed intelligence Which is our systemic approach for trying to unlock the full potential of human and AI for your organization. I Have other stuff, but would you do you want to ask me questions? Or will I step through the last bit? Yeah, I think I'll just ask you one question. I am conscious of time I suppose as a design firm I'm you know within the dock and you have these let's say consumer service companies coming in because you know They're not necessarily going to be experts in AI and that's really why they're coming to extension to the doctor to Get a better understanding of that when you're designing As well as these features and these interactions is as well as the science fiction fear of you know If we break down or you're using human terms you're using human senses see read touch Is there a fear there if you break down those barriers too much that you know, they might be accepted that there's a sense of you know The robots are kind of replacing humans and if you if you break down those barriers too far Is that a design is that interestingly the last couple of slides which unfortunately I ran out of time Is exactly in that space and what we found really useful is to and this is part of the design intelligence program We're moving away from purely the interaction and we're taking a much more systemic approach So a whole area of systems design where we look not just us at the AI and the person interacting with it But actually look at overall systems to understand flows of products or information To understand where the opportunities for the introduction of either automation or AI and systems are but to also understand the upstream and downstream Providers or people who are affected by that change and it's really important to get that understanding of how it fits in the overall system And on top of that we're finding that by and large that's generating alternative Roles alternative jobs or extending roles that are in existence at the moment. So workforce Replacement was definitely I think in the early days how AI was being sold But I think what we're seeing more and more now is that when we talk about these systems sustaining over longer periods of time We need to build a full ecosystem around it We need to allow for that human glue to stay in those systems because systems themselves change and evolve over time Okay, very good. So really a complimentary support to the humans. Very interesting. Thanks. Thank you very much