At the start of 2024, Aspen Digital began to explore the many challenges older adults face in a world becoming far more technologically advanced than the one they were familiar with for the majority of their lives and the challenges companies, civil society, and policymakers face while trying to ensure this population isn’t left behind.
As part of this work, Aspen Digital, in collaboration with The SCAN Foundation, brought together a group of leaders from across tech, civil society, academia, and public policy for a virtual roundtable discussion on how older adults should be represented at every stage of the innovation design cycle. This roundtable, combined with a number of 1:1 consultations with advocates and tech companies, resulted in last week’s publication of the Older Adults & Digital Equity Playbook.
Throughout our roundtable discussion, it was clear that there is a significant amount of catching up needed before older adults are adequately and thoughtfully integrated in the development and deployment of new technologies, particularly those that rely heavily on the use of human data. It’s not just industry alone that needs to step up. Civil society and public policy have an obligation to strive for increased digital equity of older adults as well as ensuring substantial risk-prevention guardrails are in place.
Here are some highlights from the convening:
Fireside Chat
To help center the overall conversation, our convening kicked off with a fireside chat between Anika Heavener, Vice President of Investments & Innovation, The SCAN Foundation and Kara Carter, Senior Vice President, Strategy and Programs, Staff, California Health Care Foundation. You can watch their conversation in the video below.
Watch session on A.I. & Health Equity
[Music]you are listening to a conversationhosted by Aspen digital a policy programof the Aspen Institute in collaborationwith the scan Foundation the olderadults in digital Equity work broughttogether a group of leaders from acrossTech Civil Society Academia and publicpolicy for a virtual Roundtablediscussion on how older adults should berep presented at every stage of theInnovation design cycle to help centerthe overall conversation the conveningkicked off with the following firesidechat between anuka heaver the vicepresident of Investments and Innovationat the scan foundation and Cara Cartersenior vice president of strategyprograms and staff at CaliforniaHealthcare Foundation you can hear theirconversation shortly and to read ThePlaybook that was produced as a resultplease visit ourwebsite okay so let’s dive in so we’rehere to talk about older adults andtechnology which I’m um honestly excitedto learn about as much as I am to talkabout and I will say I think for mostpeople myself included older adultsaren’t necessarily the first segmentthat we think about when we think abouttech Innovation from the scan foundationand from your purch tell me a little bitabout this demographic and why the techindustry should be thinking about thepopulation yeah first and foremost Ithink a lot of of us on the call areaware of the broader TR the Aging babyboers but both globally and in the USbut also the magnitude of this I thinksometimes the numbers we’re not aware ofand like for over and old 65 Plus inAmerica this is projected to increasefrom million in 2022 to 82 Million by2050 and so that’s a huge in terms of ademographic Shi and all that entailsfrom purchasing power to how they willcontribute and participate in societybut that also if we think about thetooling that we will use to live ourlives it’s the demographic that thosetools L AI automation Advanced analyticswhether that’s showing up in healthcareor the way we live our life it has toserve this aging population noting alsothat this is a population that has seendifferent seasons of TechnologyEvolution the old adult populationdefinitely lived in a world where theywere so ready with ped paper and forsome of us all the call we still had tolearn cursive in elementary school andso recognizing that they are not thequote unquote digital natives doesn’tmean that there’s not digital literacybut they have a different generationalstory around how they have utilizedtechnology we see an incredible Marketopportunity but also a different way ofapproaching technology that definitelyshould be looked at from a consumerperspective but also C for both you andI are organization BC ofal Healthcareand so from a patient perspective aswell and so hopefully car sure this sortof in terms of if we think about thedynamic of tech being built for olderadults especially in healthcare whetherthat medication management or careforation social connectivity there’s alot behind the scenes that Health te andmuch of it has the stars and the specialeffects right now be AI enabled and thatimpacting liable ORS you and your teamjust had an incredible webinar on AIgood or bad for Health Equity comingthat conversation can you give us aquick snapshot one why you felt that wehave that conversation but what weresome of your key takeaways sure happy todo that it’s funny um why we felt theneed to have that conversation uh Ihardly I almost feel like does it needto be said I don’t know I one of thegreat privileges of my job is that Ispend a lot of time traveling aroundCalifornia getting to meet with leadersin the healthcare space so people wholead plans people who lead providerspeople who work in digital Health techpeople who work in state government andour are um policy leaders and is not asingle conversation over the last yearno matter what was on the agenda or whatthe topic of the conversation was meantto be that didn’t end up in Ai and it itis clear so everybody is talking aboutit it is someone who is rapidly becomingan older adult a solid gen xer here mywhole life has been looking at likewhere is this when I was little we weregoing to be the Jetson at this pointright we’re not the Jetson but this isthe technology that is really has thepotential to have that kind of impact onall of us and so everybody is talkingabout it and trying to figure out whatit means and at the same time there’s umthere can be a pretty big disconnectbetween folks who are BuildingTechnology folks who are deep intechnology and folks who are makingpolicy and Regulatory environment thatwill either enable that technology tothrive or create barriers so we see ourspace in the world is really trying tobring those worlds together trying toeducate policy makers and trying tocreate a space where we can have an openand honest dialogue about where we arein terms of the where do I land on theanswer I was on management consultantfor a really long time before I did thisjob so I’m going to give you theunsatisfactory management consulting itdepends answer largely I would putmyself in a cautiously optimistic bucketthere is so much good that thistechnology can do in healthcare I thinkthat is both hard to quantify andundeniable everything from if you lookacross the whole value chain ofhealthcare you start at back officethere is no health system or plan thatis not already putting ai ai enabledtechnology into place whether that’s innote taking or coding or revenue cyclemanagement or patient segmentation ormarketing and then we’re hearing a lotabout the use cases that are shinierthan that right the ones that people getreally excited about but also scarierright in clinical Diagnostics and newtreatments and new patient triage all ofthat is happening and I there’s thepotential to do good is enormous anincalculable potential to have positiveimpact on our patient populationwhen you talk about older adults I askedmy first question thinking about olderadults as end users but older adults arealso the recipients of care and involvedin the ecosystem in manyways and I think you have to acknowledgethat we’re not there yet like the greatpromised future we’re not there yetwe’re still on a path there are two bigcautionary points for me one soundsreally obvious but this technology needsto work and it needs to work reliably uhand we are not there yet we just aren’tyes AI can do better than a primary carephysician at making patientDiagnostics some of thetime but not all of the time and so thatthat is just technology still indevelopment and the other is we have tobe really thoughtful everybody has seenover the last few years not ingenerative AI in Old AI big problemswith embedding historic inequities intodata sets that drive that drive futureinequities and so as we get into thisnext generation of Technology we have tobe I think very careful and verythoughtful about making sure that we’renot doing that Healthcare there are manythings you think are the use cases thatwe all have right I use my gen buddy tohelp write my tweets right butHealthcare is not a tweet it’s reallycomplicated so I guess I put myself inthe optimistic bucket but the cautiouslyoptimistic bucket and it’s less to meabout the technology and more about whatwe do withit I want to turn back to you does thatdoes that resonate with you are thereother kind of particular considerationsthat you have when you think about thisquestion about whether it’s good or badparticularly for older adults yeah alsoas a strategy management cons not goodon made the question a little bit F andsay I don’t know that’s actually eventhe right question weshould in say thing I think we need tobe prop focused we need to continue togo back to what are the buiness problemsin healthare and how are the tool ourexpanded toolbox now cleaning AI solvingthose problems I think one of the thingsthat I’m mindful of is AI for AI stakeor it seems like everything is having anAI element these days and that is onediminishing frankly the real forth powerof AI and really having it transformHealthcare in areas where it truly has aimmediate effect for a large languagemodels gen AI to really be coming toBear the way we can think aboutimproving patient care and doing some ofthe four things that we always some backto what we run any program withinHealthcare what is the clinical imp whatis the financial impact what is theimpact to the patient experience andalso the clinici experience and so I Iwant to continue to challenge whetherit’s the EHR or digital health or now aiwhat is the problem at solving becausewhat I’m concerned about is that we’regoing to spend a lot of time moneyenergy on area that frankly maybe arecurrent way working is sufficient and sohow do we make sure that AI is fit forpurpose is one of the biggest questionsI think we need to be askingourselves to the question around AI goodor bad and Health Equity I think there’sa critical thing to dig into herebecause uh H is a persistent issuewithin care delivery full stop nonghostal we see it sh in the way patientsare prioritize but also they F us with aclinical guideline and recognizing thatthis new tool coming into the toolboxhas the opportunity to minimize thatthose inequities or deid them is one ofthe things that I think we really needto be mindful of develop what I ammindful of with when it comes to healthand how we can the standards I think weneed to uphold are one ensuringtransparency and I think this is outsideof healthcare it’s not just a healthcarespecific as the standard of transparencyaround how these models how these llmsare created and who is the developerwhat country the data set that it’strained on how do we ensure that we areunderstanding the nutrition label of ofdevelopment then what the context thatit’s going into we have to make surethat these Solutions are contextuallyrelevant and that’s one of the things Isee at really promising developments forJ is getting the cultural context incour but then there’s also a massivepiece around management and monitoringhave some conversation with some of thebigger Tech players today it feels likethe last five minutes they’re going todo that bias check that last fiveminutes they’re going to check and dothat red teaming it can’t be the laststop on the journey it’s got to meetthrough well again in representation andthe way we build that model and the waywe then do the monitoring and governanceso I see a real process of evolutionabout the way we think about HealthEquity in the context of AI becauseotherwise I do see it as a big risk tofurtherhalf of the half Nots understanding therepresentation moralized populations inthese data set is going to be vitalagain not just in healthcare I’m I Iconsider the fact that like the internethas been consume right the models thatwe are looking at and whether by thetime we’re done with this conversationchat GPT 5 will be out working our waysthere it’s moving so fast and sorecognizing that one of the new areasfor Pursuit is new data andrepresentative data I see an incredibleopportunity for us in the Health Equityjury to make sure that when we areboring new data that is representativeof marginalized populations to the tableand that can really be a modeldifferentiator yes I see a lot ofpotential around Health at G band forHealth Equity but also a lot of watchouts for us to consider in health careyeah before move on I just one thingthat neither of us mentioned that I justwanted to put a pen onity we spend Ispend a lot of time talking aboutexactly the topics that you’re talkingabout how we make sure that we haverepresentative data how we make surethat we are tailoring our solutions tomeet the needs of the most thepopulations that have the biggest needsand I also sit with like this tensionbetween when I work with safety netinstitutions this tension betweenpushing back on technology like this andrecognizing that the technology ishappening and if we don’t spend time andattention to make sure that we have theability for safety that institutions toadopt it and to deploy it we will end upin a position where the wealthy whitecommunities have access to the upsideand the communities that I care about donot andso so glad you’re underscoring thatbecause regardless of the pel or anyindustry hikee Cycles right we know thatInnovation goes to the halves versus thehave knobs first those who are wellresourced with financing sell itnetworks and so I think this is one ofthe Bigg challenges we need to pose toAI if it is exponential and itsproductivity exponential and its throughhow about exponentially accessible inserving marginalized populations andexcessivly affordable for a safety netHospital who I has a minimum margin andI learned this the other day that theaverage it budget across all hospitalsin America $7 million will shoot they’renot going to be able to afford celebrateshiny new things that in the cost of ofji today and some of theing models andbest practices again upholdinggovernance monitoring quality chain andso recognizing that we for for scale toreach mariz communties of cost actor andso how do we incentivize the big playersthat are developing these tools to thinkabout equabledistribution and democratizing thesetools now because that’s promise rightthat that it can reach More For Less andso we really want to hold themaccountable I think too here Cara yeahthat’s I completely agree awesome notingthat you and I both live in breHealthcare are there lesson to belearned around inclusive and effable AIthat we can call from other veral yeah Ithink there are I do live in breathHealthcare and I would say in some wayshealthc care is like a great litmus testbecause it is so complicatedso it is so complicated and morecomplicated than many other Industriesso if you look at the things or thebarriers of the ways that Healthcare istrying to adapt technology I thinkthere’s a lot of things that otherIndustries can learn and grab on thefirst that I would go to we do a policypoll every year of Californians and weknew we know from this year it’s justcame out how mixed Californiansattitudes are towards Ai and Healthcareit’s less than 50% are enthusiasticabout and their Healthcare I thinkthat’s something we need to work on andchange and I don’t think that’s that wasa question about health but I don’tthink that’s just about health and ainformal poll of my her’s acquaintancesfriends suggests that there is a littlebit of ambivalence in our world aboutthis technology and will it bring greator will it bring harm and so the firstthing that I think that other Industriescan learn from healthcare does has avery long history of Engagement withconsumers and communities not somethingwe always get right something we can getwrong in many ways but the orientationum in healthcare around trying to moveat the speed of trust and create trustwith patients and communities I think isa little different than other IndustriesI’ve worked in the past when you thinkabout customer engagement and customervalue chain and I actually think aroundAI it’s a it is a great area to learnfrom so what can what can we do in allIndustries to make sure that we’rebuilding more consumer trust uh and moreconsumer enthusiasm for engaging withthis kind of Technology the second Iwould think about in the spaces that Iwork in healthcare we have a lot ofconversations about Labor and Workforceright now and I do not think that it’srestricted to healthcare we sat andwatched sag actor gild this summer umand we’re all watching the drama unfoldwith h Scarlettjohans and and open the eye like overthe last days these questions about howyou can create tools that I mean I Itruly believe that in healthcare theworkforce question is is asking thewrong question I think these tools cansupplement support and lift up thehealthcare Workforce rather than lookingto replace a Workforce and I thinkthat’s something that industry can bethoughtful about so we’re looking formore productivity we’re looking for morejoy we’re looking for a happier morefulfilled Workforce and that might meansome change but I don’t I I think wehave to be thoughtful about how we aresupporting our human beings notsupplanting them and I think that’sthat’s a universal for me it’s like auniversal truth so that’s those are thetwo that I would go I’d go to thinkingabout engaging customers and engagingtheir Workforce one that I’m actuallyincreasingly aspired by is cybersecurity so them learn around how weprotect our data but also how weeffectively not just protect our databut also the appropriate applicationutilization also some of the work thatis going on around AI Safety Red teamingthat’s so critical to health care andupholding the quality and standards ofcare and also the management of patientdata so I’m really excited to take apage from some of the cyber securitylearnings that I’m comeacross there are some comments andquestions in the chat that mentionedit’s even worse that halves didn’t don’teven have to pay going back to the Havesand Have Nots most fenders would bewilling to work with them slowly to getthebrand would love to hear your thoughtson that there was another one I said weshould also think about the balancebetween efficiency and Effectiveness andthe human interaction componentespecially around older adults who mightbe needing care and Assurance or arevulnerable in some aspects so those aresome additional thoughts anything youall would like to say in response toyeah I would say yes toboth agreed yes the the point on whopays and how you pay I just think is ait’s just so important not just in termsof buying technology or buying newapplications really all the way back tobasics you’re talking about being ableto afford the energy right and energycosts and you’re talking about uh evenif you have the same application I canimagine a private academic MedicalCenter will be able to run on a stackthat can run 25 million variables to doone Diagnostic and a health Center willnot have the technological stack to runmaybe 3 million variables so it’s it’sall really basic ways that I think wehave to think about how we’re resourcingthis to make sure that we don’t createan even bigger inequity going forwardand on The Human Side I totally agreethat was part of where I was going withthinking about how we’re supplanting theworkforce there’s no there’s never in mymind going to be prove me wrong in 50years I don’t know but I feel likethere’s not ever going to be areplacement for what your care providerdoes for you they look you in the eyesand they have a conversation with youand help you understand things look I’vebeen a patient I’ve been a cancerpatient many moments that I’ve been veryscared there’s no replacement for thathuman touch of someone to hold you inthose moments but I can say that whathappens when a physician’s done talkingto you they turn around and they go toup toate or they go to Google and theytry to figure out what’s going on andthere are definitely ways that AI canmake that process more seious for themless painful and honestly less jarringfor the patient yeah car I think whatyou’re underscoring there is that Ithink we’re again if we keep this pringfocused it’s going to create theopportunity for Solo’s productivity andput things that can gum uphealthare they have a smaller piece ofbrain capacity and M but that it getsback to that physician patientinteraction and again think thatconnection like we’re never going tomove away from Desiring that connectionU so ler markets are going to shift Iknow the hypothesis around airing theindustry whether that’s selfcareeducation caring Industries will bedisrupted but we’re always going to wantthat human connection and so that Ithink of the promise of AI is it’s manplus she not man ver that comes to theend of our fireside chat thank you carathank you anuka both for Le kicking usoff with a great Lively conversation andI really appreciate you all framing thisthe following conversation you touchedon so many issues in just that shortamount of time that concludes thisconversation
Anika Heavener opened the fireside chat with the observation that “older adults aren’t necessarily the first segment that we think about when we think about tech innovation.” As the Playbook explores in depth, this is particularly true as it relates to new AI-driven technologies. When 1 in 6 adults in the US were 65 years and older according to the 2020 Census, it’s hard to understand why such a significant portion of the population is not a more integral part of the development process of a new piece of technology.
Kara Carter kicked off the rest of the convening by positing: “how can we make sure that we have representative data, how can we make sure that we are tailoring our solutions to meet the needs of the most?”
Defining “Older Adults”
As the roundtable continued, one of the most prevalent questions we heard from participants is who exactly should be considered an “older adult”?
The US Census defines a senior citizen as anyone 65 years old and above, but if age is the only determinant of who is an older adult irrespective of other factors such as industry, geography, race, gender, will that data be as inclusive and useful as it could be?
In exploring what it means to be a “senior citizen,” Forbes states that “researchers agree that it’s important to manage perceptions of people of different ages and avoid ageist stereotypes.” This is particularly imperative when building new technology and how it might be used or not used by older adults. As a result, technologists need to have a more specific understanding of who the stakeholders are and intentionally look at the intersectional identities of this group.
One roundtable participant made the point that without proactively “going the extra mile to get people who are not usually represented in the data and their experiences included in the development of these tools” we won’t be able to meet the diversity and intersectionality of needs in any group of older adults.
Overcoming the Trust Deficit
Rebuilding trust in communities should be a top priority with all participants agreeing that it should start with honesty, transparency, and a real commitment to mitigating past, current, and future harms.
The lack of meaningful digital equity and inclusion in the design of new products has contributed to increasing levels of distrust among older adults and how they view the tech industry as a whole.
When additional aspects of a person’s intersectional identity, such as gender or race, are factored in, there is even more significant skepticism that a new tool or service might prove beneficial.
These trends among older adults dovetail with a dramatic drop of general public trust in AI-companies, from 50% to 35% in the last 5 years. Participants also pointed to the drop in trust among older adults exacerbated by companies’ seeming unwillingness to take ownership of mistakes that might have caused harm to communities. For example, AI-driven tools are making it far easier for older adults to fall for scams designed to convince them to send money or sensitive personal information to seemingly trusted sources.
Digital Equity Requires Involvement from Everyone
Participants in our roundtable agreed that a holistic approach is necessary to improving digital equity for older adults with important roles for tech, government, and civil society.
Civil society and community leaders can provide insight on shared definitions and additional data points around older adults and subpopulations
Older adult communities must build capacity -and be provided the resources to do so- to better engage with tech companies on a level playing field where their lived experiences and perspectives are easily received and represented in the product design process
Roundtable participants acknowledged that policymakers have been struggling to match the pace of AI and can better meet the challenge of rapid innovation by working more closely with both tech and impacted communities.
With adults aged 65 and older on track to make up 22% of the US population by 2040 (up from 17% in 2022), it’s past time for technologists to seriously incorporate them into all aspects of future innovation. There are multiple avenues that need to be taken to increase digital equity in the older adult population to be done correctly. Defining who is an older adult and re-establishing trust with the community are just two of them. With insights from the virtual roundtable and interviews with major technology companies and subject matter experts at the forefront of this work, Aspen Digital created a playbook to help technologists, civil society, and policymakers make significant strides towards increasing the digital equity of older adults in emerging technology.
Shanthi Bolla originally joined Aspen Digital as a Program Manager in fall of 2022. As Senior Program Manager, she works on special projects across the teams with a focus on empowered communities to provide project management, research and operations support. Before joining Aspen Digital, Shanthi worked for 6 years in the tech industry in various public policy roles focused on stakeholder engagement at the state and local levels after beginning her career as a community organizer. Shanthi is currently based in New York City.
{"includes":[{"object":"taxonomy","value":"135"}],"excludes":[{"object":"page","value":"204170"},{"object":"type","value":"callout"},{"object":"type","value":"form"},{"object":"type","value":"page"},{"object":"type","value":"company"},{"object":"type","value":"person"},{"object":"type","value":"press"},{"object":"type","value":"event"},{"object":"type","value":"report"},{"object":"type","value":"workstream"}],"order":[],"meta":"","rules":[],"property":"","details":["title"],"title":"Browse More Posts","description":"","columns":2,"total":4,"filters":[],"filtering":[],"abilities":[],"action":"swipe","buttons":[],"pagination":[],"search":"","className":"random","sorts":[]}