Hey there, welcome back. It looks like we are blind now. So just pick administrative announcements. First off, next week's October 21st brown bag slot was open, but we managed to fill that. Jer Thorp will be giving a talk on living in data, which should be interesting. You may remember JR was scheduled to give a talk earlier in the semester but was knocked offline due to hurricane issues. Administrative announcement, please just be sure to post any questions that you have in the chat or the Q and a. And I will try to get to those at the end of the talk. So today we're very fortunate to have with us Dr. Beth, mind that many of you, of course know best, but for those who don't, she's Regents distinguished professor in the College of Computing, and she's also an executive director, George texts Institute for people and technology. Among the many, many things that she's been involved with, she co leads the Emory Georgia Tech program to empower people with mild cognitive impairment, which I think we may hear some about today along with some other projects. Beth basically has told us that she's going to give kind of a retrospective of her work on aging in place, culminating in this really cool thing that's wrapping up. Now the AI curing Institute, which is really a big deal for Georgia Tech. It's a major NSF funded institutes looking at the future of AI in this case, the context of health and aging in place. And with that, I will turn it over to perhaps take it away. Thank you, Keith, and thrilled to see so many of you here. I can only see the attendee list if I can't see your faces. But I see that Amy ride reuse from the cognitive impairment program has joined. So she'll take all questions about CEP and scientists turnover, who is the PI of AI carrying as also join. So this is a great, They can field all of the questions and I will just try to pull all these pieces together. As Keith mentioned, really my goal today is to talk about how US Inspire research with older adults. The work that we've done at Georgia Tech for the past 20 plus years is now culminating in to informing the future of AI. And so this is really about what we've learned by working with older adults. It started when I started at Georgia Tech with the founding of the Aware home and developing our agenda for technologies for aging in place. So I'm going to talk about insights that we've learned over the years such that, you know, aging and support and health care in the home was really a team effort. It's a team sport. On that proactive care is going to require personalized sense-making. And we'll be able to draw that directly into the discussions of what we need an AI and that health and well-being as a whole is a journey. And so what Sonia and others will hopefully see that I am doing is starting to create the threads for why we have a new institute in a caring that's really looking at this personalize and longitudinal understanding and supportive older adults in their home. But much of this research has also then been informed by our current collaborations between Georgia Tech and Emory for the cognitive empowerment program. And in this case, we're working with older adults. We've shifted from healthy aging, so to speak, to older adults who have been diagnosed with mild cognitive impairment, which is an early stage that can lead to dementia and Alzheimer's. And in that in that engagement, we've learned so much about how to build trusted interventions and how to build technologies and interventions that are compensatory for those declines in cognition. And then as I hope to conclude within the time we have today about how all this will now come into this discipline for personalized, longitudinal and collaborative AI. So the teaming aspect, the longitudinal, the journey aspect, as well as the first License making all coming together. And what we hope to do with this new NSF Institute, AI Karen, this is a talk in one slide. I'll end on this slide. And everything else I'll hold to pull the pieces together around this narrative. So I started as a faculty member back in 1998 and took my experiences from Xerox parc and working and ubiquitous computing to my agenda which I called everyday computing, which was to understand how our everyday life would be transformed with having technology ubiquitously integrated into daily task and activities and routines. And very quickly that agenda became focused on interactions in the home. The easiest way to take us out of a workplace setting was to drop us into a home setting. But then the look for questions and priorities that really mattered in home environments. And so for us focusing on health, well-being, I'll families take care of each other. And then in particular, looking at the needs of older adults who wanted to age in place was a driving agenda and huge credit goes to Wendy Rogers and many others that I worked with in the early days. What was interesting for us was because we were not working in the context of medicine or traditional health care. We really did not approach this from a clinical perspective at the time, but approach it from daily decision-making and daily activities. And for that, it meant that my observations over the years and the work that we did as teens was that health was very personal. It was social and it was negotiated. It wasn't about doing what the doctor told you do, but it was about making all of those daily decisions that form the fabric of everyday life. But I'm going to talk about three projects within that. That gave us our at least my early, very early training I in this space. And some early insights and questions that I continue to carry with me today. I have worked as just what we've done at Georgia Tech to bring these questions in this agenda nationally. So at some point after being an assistant professor, I stepped up as the director. They were home Research Initiative at Georgia Tech. I was able to work nationally at the CCC around many workshops focused on aging in place, research agenda and smart health. And then most recently, the last few minutes of the Obama administration, I worked on P casts. So the Presidential Council of Advisors on Science and Technology, I'm going on so really, really proud of the report that we put out on independence technology connection and an older age. So it's been important for us at Georgia Tech not to just do good work care and to publish it, but to, to continue to be leaders in bringing that agenda nationally. And then in the case of the peak has report but also understanding the policy implications, what else needs to happen for these types of technical and design interventions on to, to make a difference. And that is Keith mentioned for the past decade, I have been the founding director of iPad, which has been another platform for us to take these questions around lifelong health and well-being. But understanding the, the future of the human technology frontier and the different types of technologies that we can bring together to make the future, to really transform the future of AJ. And i'll, I'll mention at the last second here with our pillars for iPad, I'm really interested in the future about how home lifelong health and well-being and more cities and inclusive innovation can come together. So one of the things that I hope you'll see in a carrying is we go to the later years of that institute, is that we will not just be looking at families at homes, but we'll be looking at the role of communities and supporting residents, especially for individuals who live at home alone. So what does it mean? Not just that home technologies, but the entire design of smart cities, smart community technologies and how they can come together. But that's a preview. Come back and talk about that in a few years. So, and then finally, all of this will come together within our new Institute. So what does it mean for us to understand about needs driving the future of interventions for older adults, and what does it mean for us to create an aware home? So my very first project, and some of you have seen the butterfly pitcher many times before. But we learned early on that this notion of supporting older adult aging in place was not a question of just focusing on that singular user or individual. But it meant focusing on the family and care network that surrounded them. And when we worked with chaplains who counsel families around questions of should someone continue living independently in their own home, or should they move to an assisted care setting? What we heard over and over again, Was that a major factor in this was the concern or anxiety of the larger cared network, the larger family. And so we sought out very early to design the system that we eventually called the digital family portrait to focus on this peace, peace of mind question. And for us in or late nineties, turn of the century, what we were doing was instead of focusing on technologies where something had gone wrong, right and alarm a 9, 11 type of situation, which is where most people in the field who are working. We were looking at 99.9% of the rest of the time to convey how would someone doing in their home. And in doing so, we landed up identifying a question that I think will be a query defining question for me. Which is, you know, was today a normal day? Because in the end that's what family members would as they would say, my mom lives by herself. She's doing fine, but I worry I can't be there all the time. And I would just love to know what is today a normal day for her or is there something I need to be aware of either in the moment or over in the long-term. And this question which we identified and 999 is still a question that is important for our field. What we learned from designing the digital family portrait and working least long-term with one individual family, was that technology also had another powerful that we did not anticipate. We could address through good sensors and good design this peace of mind questions such that an older son could, could check in on his mom and feel better about how she was doing. But what we learned when we interviewed or felt was that that connection, establishing that connection between sensors in one home and the visualization another actually change their social relationship. And in this case in a positive way that she felt more connected and more reassured through that connection, even though it wasn't tangibly visible in her own home. And so this notion that how technology can play into and connect to social relationships. They're not to be much more profound and like this question of what's today, a normal day as continue to carry with the carried with us in our work. Though as we move through different projects, I and their day did particularly well because you guys can look up the last time the Nokia's smartphone was in use and look like this. We then started to look at how chronic diseases associated with older adults could also impact aging in place in independent. So this is work with Lena men McKenna on our tools for diabetes management. And in this project we learn so much more about the personalization that was necessary to be successful in the tool to help someone who is newly diagnosed with diabetes. What we ended up building technologically was pretty simple. It was the ability for someone who has diabetes to, to capture a question or frustration. Take a picture, take a blood sugar reading, write something in a, in a web form. And for that information need to be asynchronously shared with the diabetes coach who could then help them through the day-to-day challenges and things that they were trying to do. Two in this case, there was no AI, there was just a human being on the other side. You suddenly had a better glimpse and understanding of what her her patients, the students and her diabetes education classes, we're dealing with this because she could see into their lives. It made a world of difference for her. But what we also notice when working with our users are newly diagnosed diabetics was how variable, how varied and diverse, their needs, their questions, and their parties were. So even though diabetes as a whole is a relatively well understood chronic disease and will be the topic of the iPad think-tank today at 330. How people approach it is highly personalized around their values. Though we had individuals who were focusing on what they can eat while preparing meals to their family. We had individuals who wanted to problem-solve, going out to a Chinese restaurant with their friends every Tuesday night. They weren't trying to solve diabetes as a whole. And they were certainly not trying to be the perfect diabetic because that doesn't exist. But they were trying to hone their nutrition and meals and management strategies around what they wanted to accomplish in everyday life. We also learn that we asked our users who did great things with these technologies. They, they improve their self efficacy and locus of control and they sought to make meaningful changes in diabetes management practices. So from a tool perspective, it was successful. But when we asked them about it at the end of the study and just to throw away question like, how would you describe this tool to another person who might use it in the future? They just kind of looked at us like we were crazy and said all but I was just talking to so and so the diabetes educator, so the smart phone and the secure web service and all the technology had really disappeared into that trusted relationship. Technology enabled the gathering of new data and the technology enabled that data gathering to be personalized to a person's actions. But in the, in the back and forth relationship was the dominant model for what was going on in the system. And the technology was just an enabler of that. And then finally, moving on to a third project. And the, the designing the technologies get better. Now talking about my, uh, Jacob's work with technologies to support breast cancer patients, newly diagnosed breast cancer patients under treatment. What we learned in that project was we had the same things that we had seen in the diabetes project, you know, highly personalized, focused on creating an interface that would let people drive the agenda, let people ask the questions that they, they needed, that they needed to ask. But we learn two more important things. One was that what we've seen and diabetes and we really embraced by doing our work around cancer, was that we had to have a holistic perspective. If we design the tool that was just about breast cancer. We would not have seen longitudinal long-term use because there are some days when you need to worry about chemo or you need to worry about treatments. And then there's other days if you're just getting by and paying attention to all the other things that matter in life. And we needed to create a tool that allow us to integrate those types of questions. Though it was a tool that on the tablet you can also download games and fun things. So that kept the tablet, charged it and kept it in on their bodies. I kept it with them. And then the other thing we learned was questions about how do I talk to my family about having cancer? Or how do I organize the things that need to get done around the house? Because I'm going to feel lousy during chemo. How do I kinda make sense of what I'm experiencing psychologically, not just physically during cancer treatment. We needed to design a tool that allow people to embrace and work through across all of those topics. And then the other thing we needed to do was to help people manage that experience across time. When we were working with patients, they would come home with is huge binder of information. And just imagine your newly diagnosed with cancer. You're going to start undergoing treatment in a few weeks. And you've been given something that looks like the size of the college textbook worth of information. It's overwhelming. People flipped through it at first when they're anxious and then it gets set aside and then they just try to manage the journey day in and day out. What we were able to show was that if we, if we promoted different types of information based on the different phases of the journey, preparing for surgery, preparing for radiation, preparing for chemo, alongside all of those other aspects of everyday life than people with stick with us. And they would stick with us through the journey. And then that way the information they needed was available at the time they needed it, as opposed to trying to make sense of it all at the same time. Though, just like we had seen with, with diabetes. One tries to be the perfect cancer patient. No one tries to be the perfect diabetic. They're making it through day in and day out. And we need to develop technologies that partner with them along this journey. And when we, when we, when we approach it from that design perspective, then they will stick with us. And the technologies can have the impact that they, that that's possible in terms of improving patient engagement, improving understanding what's going on, getting answers around side effects, and sticking with your treatment. Not dropping out of your treatment, not dropping out of the interventions that had the long-term goal of making you healthy. And also again, like our diabetes project, we learn that even though it was a cool tablet that had, in this case a little bit of a my mixed in in terms of suggestions, recommendations based on where you are in the journey and what you would report it. And had some really nicely designed check in the middle. But in the grand scheme of things, the use of that technology was in that relationship between a person and in this case, the trusted oncology center that they were working with, understanding how to contextualize the technology and build it into those trusted relationships was another key to the success. All right, so I'm going to go a little faster here. But what we learned from those individual projects around the supporting for a team, supporting for personalized sense-making, supporting a holistic perspective around a journey over time. Well positioned us to then have this terrific partnership which is ongoing right now between Georgia Tech, an MRA, and in this case with a focus on empowerment for older adults with mild cognitive impairment. So as a quick aside, mild cognitive impairment is an intermediate stage that exists between healthy aging. So what we had looked at before with you aware home, but is not as pronounced as dementia or what we would how we would characterize Alzheimer's. And in this case, what happens is that mild cognitive impairment and work and have impacts across areas such as memory, language, executive function, and attention. And it can mix and match everyone's experience of it is different, just like diabetes and even just like going through cancer treatment. Even though we understand these things clinically, the lived experience, there's always variable. And about one in five people over 65 will experience mild cognitive impairment. And then up to 40 percent of those may then progress into dementia or or further into alzheimer within a five-year period. So this is a significant challenge for our population as a whole, then it's a growing challenge because of our aging population. So when you're at the stage of mild cognitive impairment, the changes are not severe enough to keep you from completing day-to-day activities, but they can impair your ability to do those m. And these are what we know is IADLs, instrumental activities of daily living. Though it's cooking and house cleaning and taking medication, transportation, laundry, shopping. Working with personal finances, all of these activities that make up a good bit of the fabric of daily life on a day-to-day basis can be impacted by mild cognitive impairment. And what we want to do within the kind of empowerment program is to understand the potential implications with our families that we were working with and then develop interventions and support for them. For those of you who come from the HCI community. This is very much a participatory design action research on framing in terms of how we work with people with MCI and their families. The focus for individuals is to identify their strengths and abilities and help them to remain as independent as possible. So very much like our aging in place agenda. But just because one person has a challenge in one area, cognition doesn't mean that carries into others. And this a will challenge and a design trade-off, which is that we really want to focus on building on strengths and only compensating when challenges arise. And so in the, in the therapeutic field, this is referred to as compensatory strategies. So if you have difficulty remembering, remembering task in different types of to-do lists is just one way of keeping track with things. If you have difficulty remembering to do something on a regular basis, like taking medication, building different types of reminder situations are the kinds of things that we're looking at. But it will not work. If you tried to do everything, then you go from empowering to taking control. So part of what we're want to understand is how do we build these interactions to technologies in a way that we're fostering self-efficacy. We're building agency, we're building confidence. And we're doing this all along side. Therapeutic lifestyle interventions such as improved exercise and nutrition, which will also help sustain cognitive capacity. So all of this has gotten pulled into the cognitive empowerment program. Ahmed is an amazing opportunity to be able to work with this, with this, with this group. I'm Amy Rodriguez, who I know is in the audience. I'll works within the therapeutic programs. Craig simmering, Jennifer Du Bose and other colleagues over in the College of Design have focused on the impact of the built environment. In terms of that also impacting cognition, social engagement, exercise, nutrition, everything. I help coli the technology group and then we're also thrilled and this will continue with our work with AI carrying to be able to continue foster new research and new innovations within the program. We have. It's a big deal. It's over 30 staff. We have had now 85 family, so 85 people with MCI and identified care partner working through the program. We opened our doors in January of 2020. So if you do the calendar math, you realize it we got to operate for about 2.5 months before the pandemic hit. So we had to also very quickly figure out how to have a virtual online version of this program alongside this. And then we've continued to work, as I'll describe shortly, with different types of technology interventions that continued to build across what we're doing therapeutically, what we're doing in the built environment. So to give you a sense that therapeutic program involves supporting different types of physical activity, yoga, dual task training, THE foster social engagement including art Explorer, mission, and includes members of people with mild cognitive impairment, bringing their own talents and skills. So one of our members helps teach ukulele. Right now. Lots of work on compensatory strategies for cognition, from calendar or training to other ways of remembering and managing information, as well as the collaboration around that information. Tremendous amount of work and functional independence and occupational therapy. So strong relationship with the home environment. How do we support home safety, fall prevention, safe driving, medication, and thought. And then as a whole, what are other things that can help with emotionally and coping with a diagnosis of NCI and improving things such as your diet. So all of this is in the mix. It is a 12 month program for people who are enrollment with the program. That also allows us to have pre and post measures across a number of factors within the program. And we are continuing to adopt folks as they graduate from the program, as alumni. And they can still continue to do research and interact with us. So we're actually going to have our, our second graduation ceremony on next week for our third, fourth cohort. So the tech team, I wanted to throw this up here because you'll recognize a number these folks. So Gary Clifford who is jointly appointed between Emory enjoy to check my column like this. But Brian Jones directed The aware home. Has a huge role in the technologies that we're creating in the home environment. And then Brad fading director of home lab, is working with us for for, you know, being able to deploy out to as many families as possible, potential check interventions. Jeremy Johnson is our fearless application developer and Selima law force works with us. I'm with Jeremy in terms of pulling together and stitching together the whole program and many, many others. So we work across technologies for mobile technologies for support. So you'll see elements of our diabetes sense-making and our cancer journey applications come into mobile technologies. In technologies churn ensures safety and well-being. To amplify the program, such as amplifying physical exercise, supporting, and creating companies compensation around different cognitive tasks all the while, while trying to measure within the home environment. The changes we get back to that question was today a normal day? Is a person sustaining their activities? Do we see trends? Do we see changes that could signify decrease cognitive capabilities, but can also signify depression are other, other factors that could be intervened. If we have an awareness of what's going on. We have an amazing app now that Jeremy and others have developed that came straight out of that breast cancer work that we did with my journey compass, and my path. But in this case, it provides continual access to the program and access to the resources. We have. Now a set of technologies for different types of sensing within the home environment. Though, if I look back to what we did with the digital family portrait, where we had very simple sensors just to measure a simple level of motion within the home. We now have kind of a plethora of ways of getting at this information. And we're able to now work on different our families on different safety at home concerns. So ignoring the cartoon picture, what we hear from our families is that there are very significant concerns around leaving the stove tough on, leaving doors open, leaving water running, something being a miss. And if we go back to those peace of mind questions that we had at the very early beginning. Understanding is, you know, is all well within the home or their difference, a different scenarios, different concerns that a person needs to attend to. And this is especially important for our families with cognitive empowerment program because we may have someone who's diagnosed with mild cognitive impairment, but their spouse may still be working. And so when they're out of the home working or even in a home office focused on their work. There's the question of is my spouse inner join, okay? Or do I need to worry about that? They haven't returned back from walking the dog or they've left the stove top on our they've turn the water on but then walked away from it. So it's not, in some sense it feels like a death by a thousand cuts. In the sense of like all the day-to-day worries that can accumulate. But if we design technologies that can boast sense these situations and provide non annoying alerts and monitoring of what's going on. We can actually change kind of that fabric of that, of that daily experience. One of the areas that we have explored with the cognitive and proud of our program that we had not previously done. So just based on, on technology suitability has been the use of conversational assistance though. Mars it would be who's been working with me on a pandemic inspired project where we've been deploying Google Home, Google Home Hub technologies into the homes of our kind of empowerment. Families. Willing at first just to see what they would use within existing commercial product. One of the things that we're doing in this translational space is just the even see how technologies that are commercially available and research technologies it is commercially available. Are they suitable and can they be useful for our families? That we have over 20 families using the Google Home Hub. Tom was already presented some of this work at the last chi. And so we have found some really interesting results. First and foremost has that we, because we've been able to provide better training and introduction of the technologies. Our older adults, even with mild cognitive impairment, have better usage statistics compared to the, compared to the commercial field. And we, we, we deeply believe and are continuing to prove that it's all about the training and the support for using the technology. So even just the initial configuration of the technologies that makes a world of difference for that initial adoption. One of the other areas that we've seen, and this is just fun, has been the use of the Google Home for adding music and adding fun and adding and enjoyment in daily life. It's important in this research that you not think of this is, is all about okay, something is wrong and I'm going to go fix it. But just how can technologies enhance daily life activities and the ability to add music and to add some level of interaction on a day to day basis can be really important for someone with mild cognitive impairment. And a number of our families talk about how they'll use the Google Home to play kind of appropriate music. Like if they're making Italian at, at dinner, they'll play Italian music while cooking. And we know from other studies that, that type of social fostering, that kind of social interaction and fun Can, can actually be quite important for sustaining cognition and social engagement when you have NCI. And then the last area has been a big aha for us and will continue into our future work. Which is how the technology access, scaffolding or support between a person with MCI and their spouse, their care partner. Though we have a number of insights coming from watching them using that where it's essentially the care partner. Care partner is setting up routine, setting up reminders, adding information into the system that can later be utilized or to the benefit of the person with MCI. We're referring to this as scaffolding. And what we've seen in the interviews is how important it can be to the care partner that the Google Home can, can take some of that load or some of that work in the work of care off the shoulders of the care partner. And so even if it's a matter of asking like what day it is or what's on the calendar that day or reminders for meatus medication. To be able to take 10 to 20 percent of those questions off of the care partner and into the Google Home. It has been pretty powerful for some of our families. And so this is an ah-ha, that, okay, if we can do this without really kind of any specialized support within the Google Home, What can we do going forward? So one of the areas of work that we've, we've accomplished so far has been to look at that vexing problem with medication management. Many a researcher has spent many, many hours trying to develop technologies to help people be more adherent and taking regular medications. In some cases, these medications are highly important. In terms of managing other, other medical, the medical ailments. We know that as a whole that older adults tend to take multiple medications on a daily basis. What we also know is that most of the approaches to medication management fail pretty rapidly. People miss alarm, say hear an alarm, but if they don't take their medications immediately, they get distracted, they go do something else. They resent the alarms and reminders. They tend to ignore them. In our case, this with people with mild cognitive impairment and even if they use kind of state of the practice, which are those little medication boxes with the label the label box is Monday, Tuesday, Wednesday, is that they may come up and say, Oh, did I take my medications? I don't think I did. And even though the Thursday box, like the box for today is empty, then they'll say, Oh, okay, it must be Friday and then they'll take the medications for Friday. So it's a challenging it seems simple to describe and it's a challenging problem to come up with a successful design. And I've been really impressed. Press was no hierarchy and canal and the rest of the teams work in creating the conversational assistant. That does the reminding. Then also circles back. Though, if the person doesn't confirm that they taking their medications, it will come back again. We're now going to integrated into the med box. So we can also see when folks have interacted with the medication box. So maybe they were reminding is not as persistent. And then it also loops the conversation. Also loops the care partner and to say that the person with MCI, if you're taking their medications though, taking some of that concern and worry off their shoulders. And so what we get, for example, is this type of engagement rate, which is just terrific to see that as we iterated on and improved that certain level of persistence but kind of software systems around checking in to see if medication been taken as opposed to telling them to take their medication. We saw more use of the of our skill in Google home. As opposed to, you know, kind of getting over that novelty MAMP and seeing less shoes. And in some of the interviews we know with the families that we're working with that this has been kind of a night or day type of type of experience where they have less anxiety as an individual, but also as a couple around medication management because the Google Home is is taking some of that load off of their shoulders. All right, So what we've started to see in this work has been about creating, creating technologies that allow for that little bit of compensatory strategy. Noticing that something's been left on, noticing that the door has been left open, checking in around medication. All of this is at a level of not automating what happens within the home, but providing just enough support around the edges as a compensatory strategy. And then the other thing that we've been able to see is again, this understanding of where trust goes into this. So part of the trust has been around the fact that these interventions are built inside a larger program, program with Emory and around empowerment. So again, this is not the same thing as we're just going to put an app on the App Store and, and encourage people to download it. This is built into a set of other relationships and the set of other commitments that allows the technology to get a foothold. The other thing that we've heard is that because they are under design to the point of being relatively simple but reliable, that they are they are reliable enough that our families start to trust them. So that plus kinda the interaction with the Google Home, for example, as we've heard comments like, oh yeah, you know, it's just like the Google Home is another member of our family. And that's in some sense the highest R that we can aspire to, which is that the technologies become adopted into that care network. I think that's what they mean by another family member and that there's enough trust there. So what are we gonna do with this going forward? As we now work to create an NSF Institute for AI carry the motivation, the US Inspire in the societal. A motivation for the institute is still very much the same as where we started with aging in place. Back at the beginning of the ware home. We're looking at a growing population of older adults, were looking and understanding that older adults will, you know, what will come with that is generally multiple chronic diseases, other forms of impairment, things that we would label as disability. And that each individual is managing that within their family and within the care network the best they can. But no situation is identical. And that within that, the role of caregiving and the cost of care giving is quite high. So anything that technology can do to become a member of that care network can help offset those costs and start to make a real difference in people's lives. The AI Institute is defined as creating a discipline around longitudinal and collaborative AI. And so what will we mean by that? Well, the first is the types of things I've been talking about thus far. The development of interactive and intelligent systems. Though, like the different, I like the diabetes app and my path and my CEP, the types of mobile technologies. Instead of having kind of very, very lightweight AI, being able to have a more robust understanding. We'll get to that in a second. And then interacting with technology such as, such as the Google Home to be able to use technologies for sensing. So this is the infant sleep sensor and the door sensor and the types of fabric is something that I showed with you and that we've explored in the aware home. And then up to you and getting to Interactive Robotics in the future. And the collaborative is that it is embedded within the community of users. So I've talk about a person and their spouse and their family members are aspirations with AI carry will be even larger teams. So extended family members as well as friends and neighbors and then the different healthcare, healthcare providers. Just like the way that we interact with the therapeutic core and the different interventions that are provided in the cognitive empowerment program. And we want to do this over extended periods of time. So not, not days, not very simple AI working within the realm of minutes and hours. But what does it mean when the system can have an understanding of your values in your routines and your habits over weeks, months, and even years. So I presented briefly to the cognitive empowerment program a few weeks ago about the kinds of things that we're bringing into. Bringing into this work. Support for routine task, support for the care teams for that types of collaboration. We've seen a lot of challenges in working with the Google Home around speech recognition. And we know this is true in the field as well. So creating systems that are more robust, the different types of speech that older adults utilize, as opposed to 20-year-old developers of speech recognition technologies. So there's a, there's a great deal that we couldn't do that. Technologies can be more robust for our families. And then moving even further is I talk to you about the importance around the priorities and the personalization that we saw on our diabetes work and we saw in our cancer work. These types of priorities will have to be recognized and built into our interactive AI systems. Again, no one is the perfect diabetic, No one is the perfect cancer patient. No one is going to be the perfect person with mild cognitive impairment. Each individual and each family is making a whole set of individual choices based on their priorities and values for the day. How do we build AI systems that are not operationalized towards a perfect norm of a task, but our operationalize to you, okay, well, this is what really matters when we make dinner or this is what really matters for medication, or this is what really matters for nutrition and exercise in this family. And then build towards that as again, as opposed to some abstract norm. And going forward what we hope is that the, the longitudinal understanding of our families is these systems hopefully become part of that care network, is that they can also be a partner in helping make difficult decisions. What does it mean to come up to an AI system and say, all right, my spouse who has mild cognitive impairment, you know, things seem to be getting worse in a do we need to consider major changes in our home or our lifestyle or consider moving to an assisted living center. These are very difficult questions to answer, and so it's quiet, aspirational, but the goal is what would it mean for that AI member of the care network to be able to be part of that dialogue and to be able to say, this is what I've seen and this is, you know, this is promising and this is, this is discouraging. But here are things that can be done to be able to work through that data. But work through it in a way that helps a family begin to make and work through those types of decisions. So that's really kind of the highest bar. For us. We're going to use all sorts of traditional and participatory use inspired research methods. What's exciting for us is that we're able to work through not only with the cognitive empowerment program, but we have collaborators and Oregon who have worked within hundreds of homes gathering data about what normal looks like in those homes. And then we're also going to be partnering with the Rehabilitation Engineering Research Centers. Both your Georgia check in at CMU that all that work with older adults but with different types of disabilities or impairments are priorities. And that will allow us, again to understand the diversity of needs that these types of systems should be able to work with and be able to respond to those diverse needs, values and priorities. Sign is going to give a talk and the rest of us are going to be speaking about the future of the Eye Institute next week. But tremendous research. And what is fantastic for us is all of the all of the experts that are coming together to work within this AI Institute. You'll see many names that you know, others that you're going to learn about. But a tremendous talent coming together with these challenges in front of us. And we are spread across Georgia Tech, Carnegie Mellon, UMass long Oregon Health Science University in Oregon State University. And that collaboration is then also working across a network of partners, both in terms of US Inspire and also in terms of IT technology and AI, as well as our education and broadening participation partners. So just fantastic to bring all of these folks to the table. And as f is very excited, this is part of their national map for the AI institutes as a whole. And again, please be looking for the advertisements for the kickoff presentation about the institute, which will be just next week, Wednesday, October 20th. So as I close and as Keith is going to remind me that I'm out of time, this is really in some sense the money slide. With respect to what I think we've learned over 20 years. We've learned that care is a team sport. It requires coordination across multiple stakeholders. And technology can play a role in terms of shifting the demand and load within that network. That you can establish connections through technology like we did with the digital family portrait. But for the most part, these technologies are trusted because they sit within trusted relationships. And that's something powerful for us to understand both how AI systems can be part of those trusted relationships, but also be embedded within the larger trust relationships. That is really how we, we make sense of day-to-day life. That personalization, which is a key, key goal for AI systems, will have to be driven by the needs and capabilities of the individual use. Families come into these systems that no one is. Is focused on being perfect or having a 100 percent compliance or a 100 percent knowledge. But they are driving the types of detective work or the sense-making based on their own values. And anything that we've seen that as goal setting, that looks like old study. They're not abstract goals. They've, they've driven by values, they're driven by the, I want to go have Chinese food with my friends every Tuesday night. And how do I integrate that into my diet as a, as a diabetic? That's the kind of value centric goals that we should be working towards. And that all of this is across the journey. And I think something that will be so profound about the center is that it's not, again, focusing on AI and interaction in the here and now, but focused on supporting people over time. We know with the breast cancer work how that can make such a fundamental difference. But we're going to build those relationships over time. To do that. Empowerment is about creating new human capabilities and celebrating those capabilities are medication skill does a little like hurrah at the end and our folks kinda laugh at it, but they also gotten to like it. So this can, this notion of that we're not trying to automate or go from a deficit perspective. We're trying to build new capabilities, even as folks are having to compensate and do things in new ways because of cognitive decline. And that if we create that fabric, that reliable fabric, we create a new piece of mind through that secure interaction with technology on a daily basis. As much as we love technology, I'm in the School of interactive computing technologies, not the point for our families. It's, it's about having a coherent design that simple and easy to use nodal babble lot of features. And in many ways we have to challenge tech conventions on a day-to-day basis. We have to aspire for things or it can be usable by a large number of people. And that our age proof. And what I mean by that is that there's so many like features and whiz bang UI conventions and the desire to keep updating things every three months that just pull out the rug of our of our families. We need to have a technology strategy that allows for those long-term relationships and allows for that. That's your sense of peace of mind. And if we keep changing things and we keep having conventions for, for user interfaces, that kind of defy anyone older than 30 years old. It's not going to work for for our needs. So we're going to have to push back from a policy perspective Even on how we want these technologies to work when they got commercial feel. So I've talked a lot, Thank you so much. There's some great questions and comments and I really appreciate the opportunity to speak with you today. Thanks, Prof. Gray. Talk. So we are officially over time, but I want to get to one or two questions any way. And so I just asked you to try to address them briefly for chug. So Anthony asked, how do you see scaling the program among care networks are truly support your model of a fabric approach among the cure community. By general observation of the medical practice was pretty pessimistic though it's fragmented, bureaucratic. It seems a big lift to get this working without such a dedicated said. No, thank you, Anthony. And that's why the kinds of partnerships like the one we have with Emory is so important. So Alan Levy use the, the head of this for us, is intent on us figuring out how to scale this across the entire Alzheimer Research Disease Center at Emory. That's huge in terms of that volume. And then the point is then to be able to create a proof of that, that they can go to other areas. Even though I spent my first decade in this area, kind of working bottom up away from the medical field. It's going to have to be through partnership for us to get there. I do worry about technology only endeavors into this space because I do not see how those trust your relationships will be able to be formed. One of the things that we've noticed, we especially notice with our breast cancer work, for example, is that technology can also be pretty good at exposing the fragmentation within the medical field. So for example, some of them are breast cancer patients because they had multiple chronic diseases. They would get one thing from their oncology doc and they would get another thing from their diabetes doc, Like how much water to drink every day. And then they would come to our tablet and say, Okay, you tell us the answer. So we're going to have to we're going to have to work with aggressive partners like Emory and others that are willing to have this patient-centered approach that realizes that you have that again, had this holistic perspective. You have to have this patient-centered perspective and to bring these pieces and parts. Our hope is that through coordinated care with technology, we can play that well on the medical side, just like we play it on the familia side. Picks but I've gotta get it. One more question. Ryan asks, perhaps you should, perhaps you could discuss a bit more on sort of the difficulties you failed and adapt and consumer technologies for these applications. And you feel the sensors, the technologies are reliable enough yet, or is there more work to do? But Google is like another member of the family. He sounds very promising. Yes, So, so, so challenging. So we probably been more time saying no to putting technologies into the cognitive empowerment program than the ones that we adopt. Because they're too complex, too difficult to use, don't know how to charge it. We also spend a lot of hand-holding in training and initial configuration of the technologies. This helps gives us more proven evidence to why those steps are so important, because we can show things that are grammatically adopted much better in our program than they are in the commercial world. So we're creating this proof points, but they're there just so many failure modes right now. So the tech industry is going to have to get serious about what it means to really work within the needs of this age group in these kinds of challenges. You know, the good news is that we have lots of evidence to show them what needs to happen. I even just a simple thing, the speech recognition. Speech recognition does not work as well for older adults. Your voice changes. It certainly does not work well for older adults who hadn't been kind of trained up to how to talk to things. Are folks tend to talk and really long sentences that can kinda ramble. And the speech recognition algorithms are about like grabbing the first little bit of it and then acting on that information. And they get very confusing responses. And then they certainly don't know how to work with someone who may be confused because they have dementia or Wildfire of impairment. Though even those foundational technologies, you know, we've, we've kinda declared that speech recognition is a solved problem. It's not solved at all. Once you start bringing in these very well need. So tech has gone a long way to go. But, you know, but, but that's why I'm excited with this NSF Institute is where able to have these partnerships and be able to create new technologies that show how it should work and then tell the rest of the world. Right? Thank you so much. This was really amazing talk. We are overtime. So I'm afraid we're going to have two indices here. There are a few more questions I know in the chat, I would definitely encourage you to follow up directly with this if you would like her to answer those. But in the meantime, let's take that. And I will see you all next week.