Webinar: Creative AI for Inclusive Innovation

Great to see people from all over the world. Austria, Singapore, Belgium, Canada. Milan, Portugal. Excellent. Okay, so I think we'll get. We'll get started now. So I want to welcome everyone to our webinar on AI for inclusive Innovation. I'm Geoff Gibbons. I'm the Managing Director for the Americas at Board of Innovation, a global innovation consultancy. I'm joined today by my colleagues Natalia Diaz and Natasha Naya, to discuss how AI can be used in a very deliberate way to actually improve inclusive innovation. This is a topic that's very important to all of us, and we're very excited to have you here today for this webinar. We're going to keep it interactive. We'll have some moments for questions, there'll be a quiz and plenty of time for questions towards the end as well. And so what I thought I'd do is just, I'll start with a brief introduction to myself and then also the topic at hand around AI for inclusive innovation. So, as I mentioned, I'm Jeff Gibbons. I've worked in consulting in innovation consulting for about 15 years across lots of different industries. And I really see that innovation is tough, especially in established companies. I have a personal belief that established companies are actually the best source of innovation and the best hope for actually tackling big societal problems rather than startups. And I've always, in my career wanted to push to do innovation better. And one of the ways to do innovation better that I've seen is actually, in a way that's actually not only more successful, but also fairer for society, is to pursue what's often called inclusive innovation. And so we're going to be talking about what that is and why we at Board of Innovation think AI can be, if used deliberately, a positive enabler of it. And so we're going to get started into the topic right now. So I think the first thing to emphasize is that creative AI, which is really a way of using AI for innovation, is here to stay. We've already seen over the last six months, at least a variety of different ways in which AI has been used to reinvent the way that innovation projects are run, but also imagine entirely new types of innovations that simply weren't possible or even imaginable even six or nine months ago. As we think about the impacts of AI. Obviously, we all know that these generative AI tools have exploded into the public consciousness. So ChatGPT took only five days to reach a million users. Where you compare that to the iPhone, which was nearly two years. And while these tools and these models are actually in their infancy still, they're already having a very real impact on the economy. And that's only going to continue. And I think the thing to keep in mind is that AI is literally the worst it will ever be at this stage. We haven't seen its full potential yet. In terms of its potential, we see that from a variety of different studies that AI will actually result in the biggest productivity increase ever in history, potentially. So compare it to the advent of the Internet, steam power, electricity, which revolutionized our society. AI is actually estimated, maybe by some kind of AI bot, potentially to actually have a much bigger productivity increase in terms of the way that people work. And if we look at creative work, which is our focus, it's already supercharging how we can actually do that creative work, both in terms of actually coming up with ideas, doing research, developing designs, even actually building new products using, for example, the GitHub Copilot platform, that massively increases the productivity of software developers, even video games that are completely created by AI. So that whole process of creation is really being disrupted. And all that's very exciting and it's easy to get carried away. But one thing we've been grappling with at Board of Innovation is that AI is actually typically seen as a barrier to inclusion. So we've started to find ways and been searching for those in which it can actually be used to improve inclusive innovation, to make it easier, better, cheaper, faster, to actually pursue the goals that we know are really important. So I'm going to hand over to my colleague Natalia now, who's going to actually take us forward in terms of actually defining what we mean by inclusive innovation. And some examples of that.

Absolutely. Thanks, Jeff. I'm Natalia Diaz. I lead healthcare for Board of Innovation in the Americas. I've been in innovation consulting for a number of years. Very, very happy to be here with you. The first thing to align on here is what we mean by inclusive innovation. And we describe it as the practice of creating products and services that are accessible and usable by as many people as possible, regardless of age, gender, ability, background, or where they are in the world. Go to the next slide. So we're going to start with a little bit of a quiz. We have here some innovations. Audiobooks, automatic doors, closed captioning, and text messaging. And we want to know from you, which ones do you think were originally developed for an underrepresented community? And so either go to slido.com and use this code, or pull up your phone and use the QR code to answer this question. We'll give you folks a couple of minutes. You can choose multiple answers here. Which ones do you think were originally developed for an underrepresented community?

All right, got an early lead for audiobooks.

Yeah, popular answer. Closed captioning. Oh, okay, nice. Audiobooks is still in the lead. Fantastic. Text messaging. Not the favorite here. Cool. All right, maybe a few more seconds for people to vote. Fantastic. Well, there's good news and there's bad news. The good news is whoever said audiobooks is correct, but also everyone else, these are actually all innovations that were originally developed for an underrepresented community. So a bit of a trick question. Inclusive innovation, as we have now seen in these examples, lead to better products and services for everyone who uses them. And so audiobooks were initially developed for people with visual impairments to be able to read books. And today it's something that allows us all to use books on the go at whatever time is most convenient. So really improving the lives of everyone. Automatic doors. Right. They were originally developed for people with mobility impairments, and now they're stable in buildings and even modes of transportation all over the world. Closed captioning was developed for deaf and hard of hearing folks, and now it is widely used to just improve the viewership experience in general with this text based representation of audio content. And the last one here is text messaging. We use it all the time. But it was originally developed with people with hearing impairments in mind. Now it's this widespread mode of communication that is super convenient for billions of people around the world. So as you can see, inclusive innovation can lead to better products and services for everyone. But there are very real barriers for innovators to come up and create and scale these solutions. Some of the key limitations or barriers for creating these solutions are ones that probably sound really familiar to the folks on this call. There's human bias that is inherent to a lot of people. This lack of understanding, understanding or empathy for the needs, the plight, the experiences of marginalized people. There's limited funding, and so there's the struggle to secure funding necessary to reach and create solutions to underserved communities. There's this myth that if you define it or design a solution for an underserved community, it won't be broadly appealing or profitable. And so it's really hard to overcome those biases and those myths to get that funding that you need to. There's also time constraints. So balancing the urgency of these needs with the constrained funding, with these longer development cycles can be a bit of a challenge. And so all these three things act combinatorially. And if left unchecked, they really can amplify each other in this vicious cycle which perpetuates this broken system. So lack of time and resources leading to less inclusive or more exclusive inclusive solutions, which continues to perpetuate marginalization, which continues to amplify that lack of empathy and human bias which makes it harder to make time and find resources to create these solutions. And so again, this is a vicious cycle that if unbroken, it really perpetuates itself. And so what we are talking about here, if we go to the next slide, is in the age of creative AI, right? In this age where AI is being used to democratize creativity, where people are no longer limited by their ability to create, it's more important than ever that we understand the limitations of this powerful tool, but also its potential in driving inclusive innovation. And so if we go to the next slide here, we now know from our prior experience that AI algorithms are made in the image of the people, the imperfect people who build them and the data sets on which they are trained. And in the worst cases, they actually end up amplifying human bias. So we have a few quite disturbing examples of this happening in action. There's the Compass Recidivism algorithm. It was a algorithm developed for policing used in the US to try to anticipate recidivism and was found because of the biased data set that it was used to train on, that it was extremely biased against black defendants, leading to disproportionately, disproportionately severe sentencing for them compared to their white counterparts. Imagine if this tool had actually been implemented at large scale, really severe consequences. The Amazon hiring algorithm is another example. Amazon created an AI powered hiring tool that was found to be biased against women as it learned to favor men after being trained on a decade's worth of CVs. It was even biased against women leaning activities like women's chess club. The Google Photos algorithm, very, very interesting tool that was used image recognition to label photos and it erroneously labeled two black individuals as animals due to the significant lack of racial diversity in the training data that was used that was primarily featuring lighter skinned individuals. And lastly, a really famous example here, the study of stable diffusion found this AI tool actually amplified racially gendered images, racially and gendered biased images. It would associate lighter skin tones and men with higher paying jobs and darker skin tone with low paying jobs and actually certain crimes. As we can see, there is a limitation to what these tools can do if we're not careful. If we go to the next page because of these landmark examples, because the data that these algorithms are trained on and the people who are building them are imperfect. The discourse is rightfully centered around trying to use AI responsibly. But there is an untapped opportunity here to think about AI as a tool to be wielded to actively promote inclusion. So go to the next slide. Great. So as we know, humans and AI alone, as we've seen in the past, can't ensure inclusive innovation, but together we might be able to. And so I'll hand it over to Tasha to take us through specifically how we can do that through the integration process.

Awesome. Thank you, Natalia. Hi everyone, I am Natasha. I'm a senior business designer, designer at Board of Innovation. And as Natalia and Jeff had mentioned, inclusive innovation, especially powered by AI, can be a tool to help us promote more inclusivity within the innovation process, really in kind of two ways. It helps us understand or helps us change how we innovate inherently, allowing us to kind of create cheaper, faster and incorporate more efficient inclusive innovation methods into our processes. That then enables us to determine and create a new set of kind of innovations and products and services that would have a better outcome for society as a whole. So creative AI is changing the innovation process as I mentioned, and we really have this opportunity to reimagine the process while making inclusion an integral part of it. So on this slide you'll find the innovation journey. AI will allow us to explore breadth and depth and in other words, it allows us to explore more possibilities more deeply. And in doing so, AI powered innovation, AI powered, kind of inclusive innovation can be part of kind of this toolkit and this process that we have more inherently. And, and so that's kind of what we've done just here. We've laid out a few inclusive innovation methods that we'll dive into a little bit deeper. But before we do, I just want to lay the foundation of the innovation journey for those on the call, just so that we can all kind of context set. So starting off with the strategy side of things, strategy is really where we want to understand the landscape we're interested in the competitive movements, in the trends that's driving the industry. The inside space is really an immersion with our customers. So we really want to understand what are our customers unmet needs, what are their experiences, what are their behaviors? And you'll do certain activities here, such as conduct one on one interviews, do social listening for example. And the combination of insights and strategy really directs us to understand which challenges we want to solve. For which challenges are the most impactful and important for us to solve for ideation is where you will then kind of brainstorm solutions to these challenges. You'll then want to prototype and test iteratively these solutions during the design phase. And last but definitely not least, we then want to go into the market with a really strong method messaging strategy, for example, or campaign design. So that foundation, I'd like to take you through a few AI tools that you can use today to account for diversity, equity and inclusion. And to do so, we'll take you through a fictitious business that we created using AI. It's called Breathe Easy. It's a digital health solution designed to help people to quit smoking smoking. And this tool really offers a variety of different AI coaching a community to truly kind of aid in this goal. So taking us through the innovation process and going through some of those tools that we mentioned and starting really at the strategy phase, during this phase, we're doing a lot of secondary research which just like AI, is often biased. The first AI innovation tool or inclusive innovation tool that we wanted to introduce to you is De Biasing Research. Very simply, it's a process of using tools that we are familiar with today, such as ChatGPT or Bing AI to identify bias in research. For example, using our Breathe Easy example, we found various studies and papers on the Internet that kind of describes the difficulties faced when trying to quit smoking. We asked ChatGPT to identify which biases are prevalent in these studies or which groups or communities of people are being excluded from these researches or these study papers. And we actually found out that there is a potential gender bias, socioeconomic bias, and potentially a lack of research done with people who have mental health illnesses or challenges, which really greatly impacts the study itself. And so having identified these biases, we can then conduct additional research and really target our research that we're doing towards these communities. Or we can even use AI to predict these groups perspectives through synthetic testing. And we'll talk a bit more about those methodologies as we continue along the innovation process. So the second tool that we wanted to introduce you to is multi language research. Now this isn't a new concept to us. Multi language research really falls within the insights phase, but also in phases where you're talking to consumers or talking to stakeholders globally around the world or, or that do not have the same language that you speak. So it can be in the inside space or even in the design kind of prototyping and testing phases too. There are tools out there such as Kudu, that really make this process of multi language research, automatic, more efficient and cheaper. So it can translate not only interview guides into multiple languages, but it can train these interview guides and, and tailor these interview guides to adopt cultural nuances and dialects beyond just the language element of can even kind of conduct live interviews with real time translations, which is actually what you're seeing on the screen right now. Automatic multi language research is really impactful for innovators. It allows us to cast a truly kind of wide net and democratize feedback from diverse groups. The one thing that creative AI allows us to do is push that envelope and generate more ideas faster and with more depth. And the one way we can achieve more depth during the ideation phase is to quickly ideate multiple different ideas using AI itself. Input our Personas that we have, either synthetic Personas that we create through AI or the Personas that we've developed previous to the age of AI. Input those Personas into AI and kind of elicit predictive feedback from these Personas and use that feedback to build upon our ideas to make it more inclusive and more to make it more inclusive and more comprehensive. So for the Breathe Easy example, we actually identified a variety of underrepresented smoking groups and communities. The one example that we'll speak about today is veteran smokers. And we actually developed a synthetic Persona for veteran smokers too. We then tested the Breathe Easy idea on these veteran smokers to elicit feedback in the first person. So we generated a bunch of positive feedback and negative feedback from this veteran smoker Persona. We then use that feedback feedback to improve upon this Breathe Easy idea that we had. And the one outcome that AI produced was to train the AI coaching feature to really empathize with the different groups of people or different situations of users that would be using this platform. So going into the next phase, we have design the design phase. So innovation, like we've mentioned, is all about kind of testing and learning and truly iterating to get to your end solution to your mvp. And as we've mentioned, AI really allows you to do that quicker, faster, more efficiently. But before we want to prototype and test with real customers, because testing with real customers takes a lot of time, a lot of resources, it's pretty expensive. We, we really want to understand and make sure that we're developing prototypes that are comprehensive. We want to also understand which customers we want to interview with. And so to do that really quickly with AI, we can test our concepts that we've generated. Our prototypes that we've generated either in statement form or image form to Understand which groups of people we're unintentionally excluding from from our concepts and from our prototypes. So in the example of Breathe Easy, we realized that we unintentionally excluded those that don't have access to smartphones, those that don't speak English, those that also have various disabilities. We then asked AI to help us enhance the solution by offering, and you can do that by offering kind of a text based or automated call system. So taking any interactions that you have on an app outside of the app into more of a call based or tech based system, which basically includes folks who don't have access to Internet or a smartphone or who may be in rural areas, for example. You can also introduce multilingual support or accessibility features such as being able to adjust text and font sizes. And last but definitely not least, we have automated bias testing that you can use during your go to market phases. Automated bias testing really helps us minimize any biases that we have in the models that we create. So in the case of Breathe Easy, we can reduce any bias from a design, content, a marketing, and even the algorithms that you can create for this tool. And there's tools that actually do this already. What you're seeing on the screen right now is Google's what if tool. But there are other tools out there such as Amazon's SageMaker Clarify for example, or AI Fairness 360. And they really enable us to identify existing biases in our models, which then allows us to cater to those biases and develop diverse solutions. So before we go to an interactive session, I just wanted to kind of summarize exactly what we've heard today day. AI powered inclusive innovation tools really allows us to identify and reduce bias. It allows us to surface groups that we wouldn't have thought about previously before and allows us to kind of test and learn from a variety of different groups and by doing so create innovations that are good for the kind of society as a whole. And what's really important to kind of reiterate here, as Natalia mentioned earlier, AI isn't a substitution for real human connection. AI allows us to iterate our products and our services faster, cheaper, more efficiently. But it's a means to an end. It allows us to free up our resources so that we may go speak to underrepresented groups that we might not have had the resources to do so before. It allows us to kind of rapidly spot more insights and challenges and truly innovate for the good of society as a whole. So with that, we're going to pause for A second, we'd love to get your thoughts on which AI powered inclusive innovation methods will be the most impactful for you both in your day to day life, but then also really impactful in your, in your role. So we'll give it a few minutes for everyone to input their answers on slido.

Just I just wanted to raise those a couple of questions in the, in the chat which I meant, which I wanted to bring up. So there was a question around the synthetic Personas, are they real? I think yes, the answer is they are real or as real as we train them to be. I'd say I think what we've found is that these synthetic Personas can be developed and trained based on real world research and real data and they can be used to improve concepts before real world consumer testing even for example, to run a survey fictionally and test the results to see if it's actually going to give you the results that you need. So yeah, a lot of different uses for these synthetic Personas and I think it really comes down to the quality of the training that you put into them. It's a quality in quality out thing. There was another question around the confidentiality and disclosure of information when you're using ChatGPT or any other kind of program like Bing AI, let's say. And I think yeah, one of the things that's really important in our work as well as with our clients is actually having a deliberate approach which we'll get into in a minute around managing confidentiality and data privacy. So in all of the work that we do, it's, it's really important to actually have the data sharing and model training turned off and we're actually seeing a lot of our own clients developing or launching their own proprietary closed environment GPT platforms which, which actually enables them to use things from a confidential standpoint. Just to address a couple of questions that came up in the chat.

Awesome, thank you Jeff. I think we've gotten quite a few responses a lot in the middle in the beginning around multi language research as a tool that we can really use to enhance our work today to allow for inclusivity. We've got a bit of predictive Personas as well. And as Jamie mentioned, getting feedback is really the most important impactful so that we can iterate on our guides or our prototypes etc. Before taking them to real people and learning from them too.

I like somebody said that ideation could be flabbergastingly impacted. I love the passion in that comment. And I think as we, you know, as we talked about the debiasing and bias identification. I think obviously we know that these models have bias implicit in their development. So the key thing is actually how we as humans use them. And so the art of getting the right prompt to actually do that debiasing, I think is critical. And that's one of the areas that we've been looking at in particular as it relates to that debiasing.

And I think it's really interesting because you can use your prompts and find biases in your prompts itself.

So yeah, yeah, gets very meta very quickly. Yeah, so it looks like there's a range of interest across a few different areas. Maybe now to move forward, we'd love to talk a little bit about a few watch outs and some of these have come up in the chat as you actually start to utilize creative AI to ensure inclusive innovation. So as we talked about, it's really important first of all to avoid reinforcing biases. So as we talked about earlier, that stable diffusion example where actually not only reinforced, but amplified biases in society in the images it generated. So it's really important to scrutinize the way you're doing research, the way you're actually prompting in order to avoid reinforcing bias and to actively tackle that. The next piece which we talked a little bit about is ethical data collection and use. So obviously for us as a consultancy, we're working with a lot of Fortune 500 companies, confidential data, there's privacy concerns, legal concerns around trademarks and all sorts of proprietary information. And it's really important to have the right governance and the right tech stack to enable you to actually manage data privacy, confidentiality, but also actually have a code of ethics in terms of how you're actually using these tools. So that's one of the things we've actually adopted a code of ethics at the company in order to figure out how we can actually use them most effectively and most safely and fairly with our clients. The next few things is just making sure that we're actually addressing those accessibility barriers. So, so we actually, you know, in the actual use of AI, obviously there are accessibility barriers. And one of the things that we think is really important is actually designing solutions and ways for consumers to actually engage in innovation research that are actually inclusive themselves, even if you are using AI tools. And it actually opens up a lot of opportunities for things like, you know, using more accessible technologies like telephones, for example, as an interface rather than everything being an app, let's say. The other two last two things is to think about is obviously the consideration of socioeconomic factors. Right. So we need to actually avoid exacerbating existing inequalities. So it's really important to assess the extent to which the solutions you're actually developing are contributing to or actually combating inequality overall. Same with, in terms of, you know, the environmental impact of innovation. And then the last thing which I think is just really important with anything around AI at the moment is just this continuous learning and adaptation. So we're several months in now into this journey with a lot of the new AI tools. It's going to really require ongoing learning, deliberate adaptation and being responsive to societal needs and also what the technology can and should do. And so we have found that it's really important to stay informed about the emerging, research, the best practices and actually get user feedback actually with our, with our clients and with consumers that we're testing with to actually make sure that we're doing things in the right way. In terms of all of this, it's very exciting and lots of new possibilities we've started to think about what are some ways in which people can actually get started on this journey in terms of adopting AI powered innovation methods. One of the things we actually have just launched over the last couple of weeks actually is a one week sprint that actually helps teams to rapidly generate 40, 50 validated concepts, but also learn about the details and the day to day practices of actually using these creative AI tools on innovation projects. There's a link on here to a page that will describe a little bit about what that looks like and how that works. But this is something we've actually tried out with a few clients and had a lot of success with. Then in terms of actual practical tips to get started, we really find there's a few things that are important. So one is actually understanding the fundamentals of inclusive innovation and getting your team trained in it, developing an inclusive innovation strategy. So understanding for your business, not just in general, what actually are the barriers to inclusion in your categories, in your industries, in the geographies that you actually are most active in and defining what you actually want to change about that, starting to engage stakeholders in that within your company as well as externally, and then actually defining the metrics that are going to help you to actually know if you're actually having an impact. The last two things is developing the right stack and team to actually do this. So having the right tech stack that has the right governance approach around it, actually having a diverse team that brings together different perspectives and having the right kind of external support and Advice where it's necessary. And then the last thing is actually rethinking what you can create and having your eyes open to the idea that AI could reimagine not just the way you actually do innovation, but also the types of innovations that you can deliver and who they're for. And so for us to move forward, we'll go to some general questions, but we've got one last interactive piece here which is looking at of all the tools and concepts that we discussed. We'd love to know from you what is one thing that you actually are curious to try out with your team and what will you differently. What will you do differently as a. As a result? There's a comment in the chat around actually striking the balance of using these tools, but also fostering the right human connections, which I think is a really important point. I think what we found is it's not just about AI going off and doing its own thing autonomously. It's actually really about how can you pair the best of what AI can do and what humans can do. And we found that actually gets to the best outcomes and makes sure that you actually have that kind of debiasing check as well as a quality check on everything that's actually emerging from any of these models that you're using.

Yeah, I think we also saw a comment that was about marketing going wrong for big brands and can AI do better in this realm? And I think our point of view is that AI is not a substitution tool for having a diverse team and for doing your diligence. It actually just makes it more accessible and more efficient and multiplies the touch points at which you can do this due diligence. And so it's an additive tool, a tool that makes things more efficient and simpler and more accessible in terms of cross checking these things for inclusivity issues. But again, not a substitute for having a diverse team that can help build inclusive marketing campaigns in this example.

Yeah, I think, and I think as you apply it to the innovation process too, I think it's not just about, you know, using AI to identify what underrepresented communities might think about something. It's actually freeing up the time in your project to actually do that and meet them in their own context and meet them where they are. So a lot of, a lot of what we're seeing is the opportunity to do more kind of like gen pop consumer testing through using AI tools and from more digital lens, and then actually freeing up that time to actually do the more ethnographic work with underrepresented communities, which is the stuff that people really want to do usually, but don't have the time or budget to do because they have to cover off all of the bases of their primary target audience that they're looking at. So yeah, we've got a lot of interest around inclusion mapping. So looking at how you actually can proactively identify for your business or your project or your category, what are the biases and what are the key problems that you could look to solve. And that's we actually find that's very useful for teams to actually think about that beyond the scope of an individual project. But actually as an organization, what are the barriers to inclusion in their business and how can they actually develop a strategy around that? We've got some stuff here around prototyping with AI before potential customers. I think one of the things we haven't gotten to in depth in this, which is, but is a huge topic, is how you can actually develop prototypes much more rapidly and even develop different variants of prototypes using AI. So you can actually have personalized prototypes that are targeted towards different types of users or communities and then also seeing quite a lot of love for the debiasing of those ideas that you've ideated. So how can you actually run them through a sort of inclusion amplifier approach that could actually make them more inclusive before you actually go into consumer testing? Okay, so maybe just like to go to any questions that we have. So a couple of questions in the chat, there's a question from Zach is, is putting AI first simply putting the cart before the horse? Surely personal creativity is the starting point. Yeah, I think actually it's a great question. I think the human creativity and the AI creativity, we see that they actually interplay with one another. I don't know that there's always a right starting point. I think it's really about how can we combine. And I think we're still figuring out what's the best way to combine those, especially in a context where sometimes we're remote, sometimes we're in person together. Next question. There's a question from Adarsh around the tech stack. So yeah, the tech stack I think is really about finding the right set of tools that are useful for you, your business, but also having the right governance approach and the right licenses and the right approach to data privacy and confidentiality so that you can actually use these tools effectively and safely and fairly. I think the thing that we've seen you definitely can't do is just sign up to a ChatGPT account with your own Personal Gmail and use that for work, that's really not going to be a good idea for anyone. And so having the right tech stack and having the right governance around that is critical. There's a question around, is innovation different to creativity? I don't know, Natalia or Natasha, if you want to answer that one.

Yeah, this is. This is quite a philosophical one. The way I look at it is creativity. Creativity sometimes rarefied, like people. People think that creativity is something that is hard to develop, hard to come by. It really is problem solving, that is all creativity is looking at things from a different angle to solve a problem. Innovation at its core, it's very similar. But I feel like it's kind of the outcome. Right. Of applying creativity to the world's problems to come up with a new solution, a new innovation. I think central to both is empathy. If you're looking at things through your own lens all of the time, you're unable to either be creative effectively or innovate effectively. So lots of common themes there. Tash, anything to add?

No, I think that makes sense. I think innovation. It reminds me of kind of the general Venn diagram that we speak about, like viability, feasibility and desirability, and kind of overlaid on that is creativity. So I think there is some nuances with that, but more to add beyond what you mentioned.

Great question from Amanda. What value proposition have you found resonates with those who feel, quote unquote, they don't have the time or resources to be more inclusive in their innovation efforts? And I see your point, Amanda too, that this is sometimes an excuse that we might talk about. Right. I would say that what we've seen is actually the breadth and depth is very important in terms of that value proposition. It's obviously easy to do things quickly, but actually giving that breadth and depth of coverage around the research that you're able to do, around the testing they're able to do is actually really important. And we feel like that's. Those are kind of some of the less talked about benefits which are actually valuable for inclusion beyond just the speed with which you can actually do things. In terms of the synthetic. Synthetic Personas, yeah, we've seen them being very effective, actually. We're just starting to do some projects where we're actually testing on an AB basis between live human testing and the synthetic testing, to actually have a realistic test between the two in terms of them. But I think what we found so far is actually it's the value of having multiple validation and testing loops before you actually spend that precious time with real people actually testing a prototype and having something that is iteratively improved and identifying the weaknesses of that before you actually put it in front of a customer. We found it to be very useful for that. And then the last question that's just come up, the latest one is how do you make a large organization recognize the need to buy into the concept of inclusivity? I think there's a few strategies that I would suggest and then I'll ask Natalia and Natasha to add. Any further thoughts? I think the first one is actually to really emphasize a lot of the examples we talked about in terms of how inclusive innovations are actually better for everyone, not just for those underrepresented communities. The second argument from a sort of business case standpoint is actually that if you actually don't pursue inclusive innovation, you're literally leaving money on the table because you're narrowing the scope and potential of what innovation you're working on. So I actually find that those arguments, I think are very valuable before you even get to the question of whether or not it's the right thing to do.

Yeah, I think, Jeff, I would echo that last point, actually. It is just the right thing to do. But if you need additional proof points, there's certainly what you've said around, you know, increased reach, increased revenue potential, but there's also improved decision making. The more sort of perspectives that you have in decision making, the better decisions that you make. There's lots of data to support that point. There's increased sort of well being and productivity of your workforce. There's also it makes it easier for you to attract top talent and retain it and it improves the reputation of your company as a whole. So those are additional data points as well.

And just to add one more insight, there is if you think of the products that you're designing for, often if you are taking the most extreme user feedback and accounting for it into your design and solving for that extreme user feedback, that will essentially trickle into your gen pop, essentially. And a good example of this is OXO Grip. It's OXO Space Grips if you want to Google that or ChatGPT it. Essentially they created kitchenware for someone with arthritis and now it's used more kind of socially within the gen pop. So I think it just makes your products even better.

Yeah, there's another question here around the impact of EU regulation on generative AI and how this might impact the competitiveness of businesses in the UK and the US where those regulations aren't there. I think it's a really Interesting question. And I would say that we've seen that obviously the regulation is going to take some time to catch up with the technology. So the most recent legislation that was passed by the EU Parliament last year. No, last week, sorry. Was something that had been in the works for several years and was addressing some of the topics like predictive policing that Natalia mentioned earlier, actually. So I think it's actually going to take quite some time for any government to actually catch up in terms of actually regulating generative AI, because it is so new, which actually puts more pressure on companies and nonprofits to actually identify ways we can do things responsibly and proactively. Positively too, in terms of how it might impact the competitiveness of businesses. I mean, my, my sense is that actually careful regulation of these matters in a way that protects the privacy of people and also manages the bias of data sets, is actually going to help us all. And the EU is so large that it's going to actually force companies developing a lot of these tools to apply those principles to the development of those globally. So I actually don't think we might see necessarily a situation where you have particularly more regulated markets. I think it's. The EU is so large that it's going to impact how everything's done in the US and the UK as well. And I'm not going to get into Brexit and whether or not that was a good idea. Interesting question from, from Zach around. How do we actually need to change the work processes away from agile sprints? They're not ideal for AI development. Yeah, I think it's actually a really interesting point. I think this is something we found is that agile sprints are based on usually one or two kind of loops of build, test, learn. And I think what we found with the development of AI tools is that that iterative loop actually might happen multiple times in a day or a week, rather than being something that you actually flow through. So I think we're actually going to see a lot of change in that to a lot more fluid innovation processes as a result. That's something we're trying to figure out with our clients at the moment. Next question. Do we think that AI and. Or the need for inclusive innovation opens up new job roles and skill sets within the innovation domain? Yeah, absolutely. I mean, I think there's a whole. There's a, like, in general with these new technologies. I think obviously there's a rightful concern around job losses that result. But also I think what we find is that the job creation that happens in the new Roles with the time that's actually freed up means that the bar is raised in terms of what companies can do as far as innovation and what's possible. And that means that to compete you actually have to be adapting to those new skills. So I would say in terms of new jobs, I think there's actually going to be a lot of new jobs in the space of actually creating the right tech stack for innovation based on an AI standpoint. Also a lot of the training around usage, around ethics, around debiasing is going to be very important. I see a lot of, a lot of potential for that in the next couple of years particularly. And then as far as skill sets, I think it's really important to think about the fact that if you actually have these very technology led, AI driven approaches to all these things, obviously fluency in those tools is going to be really important from a skill set standpoint. But also actually it's going to make more important some of the human skills that we take for granted a lot of the time. So it's actually going to make it more important to actually effectively manage complex projects when you have lots more inputs and lots more outputs within that project. It's actually going to also mean that having that human act of curation and saying the AI is telling me this, but I think actually it's not that, it's this instead is going to be really important as we go forward. Any thoughts on that, Natasha or Natalia?

Yeah, we saw this before with, when big data started and data science started becoming a thing that was more and more common in companies. We started seeing roles that were supporting the infrastructure, the training, the learning around it, and a shift in those roles as well as roles like data curator. Right. Like there was, there was new opportunities to untangle all of these new inputs and outputs. I think we're going to see something similar with, with implementing AI into innovation and creativity in this way. Where to your point, Jeff, we'll, we'll need to all become quite fluent with, with this technology, its limitations and the sort of quality control aspect and curation of it as a, as a first measure for sure.

There's another question around. Will I also contribute in cybersecurity? I guess maybe the, the, the, the lazy answer would just be yes, of course. Just as I think that AI would contribute in all these different areas. I think this is a really important area for thinking about the use of AI from a positive or responsible standpoint. Obviously there's a huge opportunity within cybersecurity to increase the ability to predict threats but that has to be done in a way that's actually fair and reasonable. There's also obviously the ability, the sort of arms race, I think, between nations and actors in how you actually can adopt AI to counteract and, or hack into other systems. So I think we're going to see, I think we're probably already seeing a massive uptick in the types of levels of threats that are possible and the investment from different governments in actually protection against what are going to be effectively smarter tools to actually hack into systems and degrade them as well.

And I would say as a general principle for AI, when we're thinking about how and when to apply it, AI actually still is very limited and does quite poorly in very complex problems that are data poor. And so interestingly, let's call it plain AI, just to separate it from generative AI has already enhanced cybersecurity by helping speed up the detection and response to threats via pattern recognition. Right. So taking this large body of data and helping it sort of narrow it down to a few anomalies that are detected, but then gen AI, which is different, Right. And think of it as not just pattern recognition and sort of distilling of large amounts of data and then being more predictive, but actually creating new things. One interesting application today that, that is being explored is creating more realistic security tests to improve defenses by simulating these potential threats more faithfully for training and sort of system hardening purposes.

One last question in the chat before we wrap up talking about inclusion from Claudia. How would you say the lack of, of access or knowledge on AI technologies and applications will impact persons, companies or countries not developed on these things? I think the most important thing to understand is that AI is here, it is the worst it will ever be and is continually evolving. And I think as a company we really think that from a board of innovation standpoint, we want to be the leaders in how do we figure out how to do this effectively, how to do this positively. We have a sense that, you know, AI is not going to replace innovators or people in the process of innovation, but people who are not using AI at all, we believe will be left behind. And so the, the key thing is to actually start that learning curve of figuring out what's the most valuable way to use AI for innovation, for the best outcomes and the most successful execution as well. And I think we'll be on a rapid learning journey over the next couple of years in particular as we figure that out and actually as the technology evolves. Excellent. Well, so I want to thank everyone for their time today, people all over the world. We would love to stay in touch around the topic of AI for innovation and inclusive innovation in particular. And we've put a link in the chat to learn more about the one week Innovation Sprint Powered by AI that we've actually launched and been trying with some of our clients recently. And we look forward to hearing from you about how you've managed to start to apply some of these methods in your work. So thank you all very much and we will be following up with a recording of the webinar.