The voices sing songs, repeat well worn lines from ‘70s reruns, explain how to cook or conquer technology. More recently, they’re from a Zoom or Meet call waiting for me and the patient to connect by video for an online visit with a loved one.

And it hasn’t just been home visits. I spent a significant part of my career as a hospitalist where, once again, TVs provided background sounds and voices as we provided care. I could piece together whole episodes of popular shows during my mid-afternoon rounds!

The lesson from this is clear but often missed: People use background voices and sounds to care for themselves as they practice self-care. Voice provides company … it entertains, educates, inspires and even, at times, confounds, but it is almost always present.

Today’s voice-AI agents like Siri and Alexa are used for narrow tasks and only usable by some for self-care. We can do so much more with voice!

The power of personalized, proactive voice

By creating personalized content, we can expand the self-care model to a shared caregiving model where caregiving teams can use it for a proactive, 24/7 connection with their care recipients.

Imagine the following scenario: A 45-year-old woman is trying to take care of her 85-year-old mother who lives 3 states away. Last Christmas she gave her mom SiriusXM radio for the news and ‘50s stations.

Her mom now listens all the time. Lately though, she’s forgotten several of her medical and social appointments and her daughter’s reminder calls often go to voicemail.

Daughter - Let’s sneak some reminders into your day between Johnny Mathis songs, ok?

Mom - Ok, as long as it doesn’t interrupt the best parts!

Daughter - I promise. This way you can easily let me know that you’ll go to your appointments.

Mom - It’s a deal!

This scenario isn’t far-fetched at all.

Voice can now be and should be leveraged to help us care for those who can’t care for themselves!

If you’ve read most or all, you should now understand that our vision isn’t just an incremental improvement to today’s internet and AI agents—it’s a fundamental departure from today’s normal … a complete paradigm shift!

To summarize how and why our vision is such a leap forward, this post starts with a table that contrasts our proposed model of virtual caregiving—a proactive presence that is user-controlled and multi-user—with the legacy online model of virtual consumption—i.e., a self-service activity that is host-controlled and single-user.

A close read of the table should reveal why the current virtual world of AI + the internet won’t ever support caregivers and care recipients—regardless of how “intelligent” the AI becomes—without the platform extensions we envision.

Subscribe now

An open challenge ... and an invitation

We close this introductory series with an open challenge to readers who still doubt whether proactive voice-AI agents with humans-at-the-helm will emerge to support a new class of online services that connect and serve hundreds of millions of caregivers and those for whom they care.

Consider first whether you agree that a new generation of voice-AI agents will emerge that are capable of proactively initiating dialogues with end-users that will both enhance their lives and enable them to monitor and care for other end-users of the same service?

Or, do you believe that today’s prompt-response voice-AI agents can already do some of both and will soon do much more along these lines? Or, alternatively, perhaps you still have doubts that the technical or commercial barriers to such futuristic capabilities will ever be overcome?

You may instead be comfortable with our core premise but still skeptical about the specifics. If so, what other human-AI-internet architecture or model do you envision that will enable tomorrow’s entrepreneurs and tech innovators to:

• Proactively initiate conversations with care recipients based on their unique needs and patterns—without requiring them to remember to ask for help?

• Leverage deep, continuous knowledge of individual care recipients—their routines, preferences, health history, and recent changes—rather than relying on generic intelligence trained on millions of people who aren’t them?

• Keep human caregivers at-the-helm asynchronously—maintaining control, providing context, and exercising judgment—while an AI agent serves as their continuous presence with the care recipient?

We fail to see how today’s AI paradigm—including its push for AGI and Agentic AI (agents acting without humans)—will someday support any, let alone all, of these critical elements of caregiving and care-receiving.

Current voice-AI agents wait to be summoned. They only speak proactively if pre-prompted by an end-user (e.g., a reminder) or configured by an app owner (e.g., Alexa’s “did you know...?” prompts). They all operate on a self-service, single-user, prompt-response model where the platform owner maintains control.

Overcoming today’s barriers to virtual caregiving

Caregiving, on the other hand, requires that voice-AI agents be proactive (PVAIs) and that caregivers be able to configure them—be at-the-helm—in order to serve care recipients in an asynchronous, automated manner.

In essence, we’ve identified a structural challenge—two additions are needed to today’s AI + internet platform if it’s ever to support virtual caregiving: Proactive Voice-AI Agents (PVAIs) and User-Controlled Services (UCS).

We also can’t envision how the AI rush of today—which seems to suggest that Agentic AI will soon have knowledge of everything and be capable of any task—will produce systems that acquire and maintain the intimate knowledge that caregiving requires.

Even if an AI achieves superintelligence by training on billions of conversations, it still won’t know what YOUR 87-year-old mother needs today. Did she sleep poorly last night? Is she apt to forget an important event today? Did she take her new medication yet? Is she lonelier than usual because her daughter can’t visit this week?

Effective caregiving requires intimate, ongoing knowledge of the individual—knowledge that no generic model, however “intelligent,” can infer from training data.

This represents an epistemological challenge: the particular, contextual understanding needed for caregiving simply can’t be derived from general intelligence about millions of people who aren’t the individual being cared for.

Apps or services where control shifts from platform owners to individual users who coordinate care for others ... those don’t exist yet, do they? And AI systems that already have the up-to-date, personalized, longitudinal knowledge that caregiving requires? Those aren’t on Agentic AI’s roadmap either, to our knowledge.

But they are our vision—our passion—and we are confident they will soon come to pass.

If you can envision an alternative architecture or AI approach that will bring any or all of these capabilities to the virtual world, we’d love to hear from you and learn.

Subscribe now

If, like us, you understand why today’s AI + internet platform can’t get there without PVAIs and UCS, we invite you to follow us at www.futureofcaregiving.com as we continue building the case and the community for this critical paradigm shift that’s been hiding in plain sight!

This is the fourth and final post explaining how our proposed innovations—Proactive Voice-AI Agents (PVAIs) and User-Controlled Services (UCS)—will benefit millions of caregivers and care recipients worldwide, enabling us to use AI + the internet for the greatest good. Below we explain what some may view as the most compelling rationale of all for this new model of e-care … it would reestablish our trust in the virtual world.

The most glaring failure of the virtual world to date has been its inability to earn consumers’ trust for much more than finding and buying things or communicating with one another—asynchronously or in real-time, via text, voice, or video. While we use mobile apps and online services constantly for commerce and communications, we don’t trust either to help us care for what matters most: our loved ones, especially our young children and aging relatives.

The most widely told stories about our online experiences involve suicide rates among the young, scams targeting all ages, blatant racism and defamation, ubiquitous phishing, identity theft, institutional security breaches, and cybercrime. And the list goes on.

Regrettably, the tech oligarchy doesn’t—or simply can’t—mitigate the collateral damage caused daily by and in the virtual world they control.

Apple, for example, assures us their proprietary ecosystem will keep us safe, but only if we stay within it and figure out how to protect ourselves on our own. Hundreds of millions trust Amazon and others to deliver what they order online, but not to keep their profiles and preferences private. Similarly, trust in Facebook or Google to safeguard personal information—rather than share or sell it—is at an all-time low. Sandwich-generation caregivers, meanwhile, worry about both their aging parents and their young children falling victim to online predators of all sorts.

Imagine trusting Alexa to entertain or check on your mother—only to discover that Amazon is using her health conversations to target ads, or that a data breach exposed her personal information or medical history to others. If users don’t trust Amazon to save and reuse Alexa’s recordings of their voice, when will they ever trust Alexa to do something more important than playing music or setting timers?

Subscribe now

What we’re being asked to trust today

Consider what we’re being asked to trust in today’s online world:

(1) AI chatbots owned and operated by faceless institutions whose primary motivations are to monetize everything and maximize profits;

(2) Systems that collect and reuse personal details about us—our online activities, preferences, purchases, homes, histories, and health; and

(3) AI models and chatbots that regularly “hallucinate” or make mistakes we can’t detect or correct without specialized knowledge.

While it’s hard to determine which of these weaknesses represents the biggest barrier to trusting VAIs to care for us and our loved ones, it’s clear that none of these risks existed before consumers migrated to the virtual world.

Giving control to users …

PVAIs with UCS address these trust barriers by shifting who’s in control.

Instead of faceless institutions controlling an AI agent that monitors or advises your parent, family caregivers and care recipients—or the professionals they authorize—control it. Instead of corporations collecting and controlling health data without meaningful consent, caregivers decide what data’s collected, who can access it, and how it’s used. And instead of being forced to trust opaque algorithms, caregivers can configure what their PVAI says and when it speaks, plus set up triggers for urgent alerts.

So, what’s the bottom line?

Big Tech has given us neither the proactive monitoring tools nor the user control we need to care for our loved ones remotely in ways that only we can. The self-service voice-AI chatbots that exist today offer caregivers and care recipients no meaningful way to stay connected, virtually yet constantly.

Happily, the solution is hiding in plain sight. We can now build systems—PVAIs—where those closest to the person receiving care, be they family and friends or trusted professionals, are always at the helm.

Having more control over our shared online experience through PVAIs and UCS will enable us to entrust AI agents with our agency and help us monitor and care for those we love.

This is the third of four posts explaining why our proposed innovations—Proactive Voice-AI Agents (PVAIs) and User-Controlled Services (UCS)—will have a profound, positive impact on the nascent market for e-care—benefiting hundreds of millions worldwide and finally enabling us to use AI + the internet for the greatest good. Here, we examine how PVAIs with humans-at-the-helm enable personalization tailored to care recipients’ needs rather than service providers’ financial goals.

Subscribe now

Caring for a close relative or friend is about as personal as life gets. A loving family caregiver naturally adapts their caregiving to reflect what they know about their relative-in-need—both their medical history and recent events.

Their lifelong relationship usually means they know one another’s favorite songs, preferred blogs or books, as well as the biggest challenges they’ve each faced over time.

Consider Sarah, a working mother herself who knows her 80-year-old mom gets anxious every day around 5 pm—the time her late husband used to come home from work. Sarah could personalize a PVAI to check in then with a comforting reminder for her mom to call her daughter or listen to her favorite music.

No professional caregiver working hourly shifts could know this … and no AI algorithm could infer it from EHR data. Professional caregivers simply don’t have the time or the resources to understand a client’s status or wellbeing—what techies call “current-state data”—or to use that understanding to personalize care in this way.

As important, physicians and other clinical professionals address only a fraction of what drives our holistic health. It’s widely accepted, for example, that only 10-15% of our wellbeing is determined by the medical care we receive, while 35-40% stems from our ability—or inability—to practice healthy behaviors. Family caregivers are typically very familiar with these behaviors, while professional caregivers know little, if anything, about them.

Studies suggest that other leading contributors to our whole health include the socio-economic environment within which we live (30-35%) and our genetic or biological makeup (20-25%).^[i]

Recently, more professionals have acknowledged the importance of considering their clients’ holistic health in determining the care they provide. Indeed, the impact of these so-called “social determinants of health” (SDOH)—our behaviors, relationships, and environments—on our sense of wellbeing is now widely accepted.

However, physicians are trained to diagnose and treat only that 10-15% of our whole health driven by medical conditions, while specialists focus on specific behaviors or conditions within the larger picture. Compounding their restricted focus, healthcare professionals generally limit their care to scheduled appointments and hourly visits, and rarely if ever interact with clients between sessions.

In stark contrast, close relatives or friends often worry around-the-clock about someone for whom they’re caring. At the same time, they instinctively consider the factors that shape their loved one’s sense of wellbeing—the SDOH—when they develop and follow a personalized care plan.

The opportunity, therefore, is not to expand paid professional care to address SDOH for all clients, but instead to provide personal caregivers with the digital tools—PVAIs and UCS—to monitor and support their loved ones’ holistic needs when they can’t care for them, in person, themselves.

Whose interests does digital “personalization” really serve?

Another major difference between institution-driven and user-driven personalization of online services lies in the motivations behind them.

Professional caregivers are motivated by their calling—caring for individuals, usually by the hour, in return for compensation. Consumer app providers, meanwhile, promote their ability to tailor apps to individual users, but their primary motivation is to get more users to use their app more often and purchase more of whatever they produce and provide.

Institutional healthcare providers are also motivated to “maximize client engagement” via digital channels, though their main aim is to improve care quality and outcomes rather than to monetize eyeballs and maximize profits.

Given these institutional motivations—commercial or clinical—it’s clear that today’s “personalization” of mobile apps and online services is designed to serve providers’ goals more than users’ needs.

Consider Netflix. An “e-commerce personalization platform” called Rebuy Engine claims that Netflix’s “ability to personalize product recommendations is second to none,” citing as proof that Netflix earns over $1 billion in “customer retention revenue annually.”^[ii]

In other words, Netflix “personalizes” their service to keep us watching and paying for videos—earning a billion dollars more per year by doing so. From an AI perspective, their algorithms “know” us well enough to recommend content we’re apt to enjoy.

Ask yourself: Who benefits most when Netflix earns $1 billion annually from their industry-leading personalization while users get AI-driven recommendations about what to watch next?

Real personalized care requires personal relationships

Family and friend caregivers are not motivated to generate increased engagement or more revenue—they’re motivated by the personal, often loving, relationship they have with their care recipient. As such, they can easily determine—if they don’t already know—what their loved one wants and needs.

PVAIs can serve as the digital platform that enables family caregivers-at-the-helm to personalize ambient, voice-first apps to monitor and support their loved ones’ holistic wellbeing.

Through the same PVAIs, family caregivers will also be able to coordinate the specialized monitoring and care that only medical and behavioral professionals can provide—bringing together everyday support and specialized care in service of their loved ones’ holistic wellbeing.

In our next post, we’ll explore how this fundamental shift—from institution-controlled to user-controlled personalization—will restore our trust in digital technology and create the foundation for a caregiving revolution.

^[i] GoInvo, Determinants of Health visualization, July 2017, https://www.goinvo.com/vision/determinants-of-health/

^[ii] Rebuy Engine post, “How Netflix Uses Personalization to Drive Billions in Revenue”, February 2024, https://www.rebuyengine.com/blog/netflix

This is the second of four posts explaining why our proposed innovations—Proactive Voice-AI Agents (PVAIs) and User-Controlled Services (UCS)—will have a profound, positive impact on the nascent market for e-care—benefiting hundreds of millions worldwide and finally enabling us to use AI + the internet for the greatest good. Here, we examine why voice tech hasn’t lived up to its promise—and how PVAIs with humans-at-the-helm will unlock unlimited utility for caregivers and care recipients alike.

The vision of talking with computers using voice and conversational AI has motivated entrepreneurs and developers for 50+ years. From the Eliza experiment in 1966 to voice assistants like Wildfire in 1996, Siri in 2007, Alexa in 2014, and now ChatGPT and its competitors, innovators have tried to let users interact with computers and mobile apps naturally, just by talking. At each launch of these next-gen VAIs, industry pundits declared that this time, voice tech would transform the virtual world.

The latest book promoting this view—based on the rise of LLMs and GenAI—is The Sound of the Future: The Coming Age of Voice Technology by Tobias Dengel.

Like his predecessors, Dengel predicts that ChatGPT Voice and the like will “liberate us from these clumsy tools (i.e., keyboards, mice, and touchscreens) and return us to the innate form of communication we’ve known for thousands of years.” His enthusiasm even leads him to suggest that VAI adoption will “redefine how users interact with technology, much like the internet in the ‘90s and the smartphone in the 2000s.”^[i]

The bumpy history of the voice space, however, offers no evidence that LLM-based VAIs will rival the internet’s impact unless and until they provide more value or utility than touchscreen alternatives—and address their shortcomings.

Prioritizing utility for users over value to app owners

The long-anticipated benefit of voice-UIs has been that they can be easier and quicker to navigate than the touchscreen-UIs we use instead. Remarkably, however, despite Amazon’s massive investment in Echo speakers and Alexa since 2015—which led to $10B in losses in 2022 alone[ii]--today’s VAIs have yet to convince users to choose voice over touchscreens for many, if any, of the most popular apps.

The matrix below shows how the lack of proactive VAIs + UCS has thus far limited the utility of VAIs to relatively simple apps or small populations of users.

The most common argument for VAIs on consumer devices—the convenience of not having to find, open and navigate an app using a touchscreen—is shown at the lower left.

At home, millions of us listen to music or set cooking timers using our voice instead of a touchscreen. We value VAIs even more when driving because it’s easier to get directions via voice and much safer. Similarly, physicians can’t pilot a mobile device when they operate, but they can speak to an “AI scribe” to enter data in an electronic health record (EHR).

All these voice-first apps, however, rely on the default, prompt-response, host-controlled services (HCS) model of today’s web. This means that the primary value users derive from them doesn’t change whether they’re accessed via touchscreen or voice—only the UI does.

Subscribe now

Addressing voice tech and app market failures

On the other hand, the PVAIs suggested at the top right of the graphic—being configurable by caregivers and capable of speaking proactively to a care recipient—can and will provide unprecedented utility to both. Indeed, the fact that host-controlled apps can’t connect and serve both caregivers and their care recipients constitutes a costly failure of the market for mobile apps and online services.

When asked what specific, caregiving market failures might be addressed by PVAIs in the future, Claude.AI named three that strike us as both common and relatable:

1. “Information asymmetry” where caregivers are frustrated by incomplete, often-delayed information about their care recipient’s status or needs.

2. “Coordination costs” when care teams struggle to communicate with those they’re monitoring who don’t—often can’t—use touchscreens on mobile devices.

3. “Access inequality” when aging adults with dementia or other disabilities are excluded from digital content and apps that could benefit them most.^[i]

A seismic step-up in utility and value

We believe that offering users control of proactive, voice-AI agents would quickly overcome these market barriers.

In short, these two paradigm shifts—PVAIs and UCS—will unleash a tidal wave of application innovation, adoption and added value that will truly be as transformative as the internet itself.

^[i] “If voice-AI agents could be configured by family caregivers to speak to their care recipient in their homes, could they address any of these market failures and, if so, which ones?” and related prompts, Claude.AI, Anthropic, 23 September, 2025

^[i] Tobias Dengel, The Sound of the Future: The Coming Age of Voice Technology (Public Affairs, 2023), p. 18-19.

[ii] Jake Swearingen, 2022, “Amazon thought Alexa would be the next iPhone. Turns out it’s a ‘glorified clock radio.’” https://www.businessinsider.com/amazon-alexa-business-failure-10-bn-losses-2022-11#:~:text=The%20Colossal%20Failure%20of%20Amazon’s,of%20gadgets:%20the%20smart%20speaker.

Business Insider

Our posts last week set the stage: Caregivers are struggling, care recipients are suffering, and today’s self-service, host-controlled web offers them next to nothing. We introduced two revolutionary innovations—Proactive Voice-AI Agents (PVAIs) and User-Controlled Services (UCS)—that can and will change all that.

In this post and those to come, we’ll explore why these innovations will have a profound, positive impact on the nascent market for e-care—benefiting hundreds of millions worldwide and finally enabling all of us to use AI + the internet for the greatest good.

Subscribe now

Most readers will agree that caring for those who can’t care for themselves is an essential part of being human. At the same time, caring for someone else is often a difficult and thankless task.

Billions around the world have worked tirelessly to care for a stranger, loved one, or friend while those they’re helping have nonetheless suffered from loneliness, social isolation, and multiple chronic or acute conditions.

As we’ve described in earlier posts, today’s reactive, self-service apps can’t bridge this gap because they require the care recipient to initiate every interaction—an impossible barrier for many who need help most. Though the latest AI is capable of completing myriad tasks, it simply can’t help us care for someone else, yet.

When PVAIs are given the agency of a caregiver—serving simultaneously as their proxy and as the automated yet amiable companion of their care recipient—they will benefit both in unprecedented ways.

The graphic below shows how the future PVAI + UCS web model differs from the legacy consumer model—by simultaneously connecting, serving and giving control to both caregivers and care recipients.

Innovations for the good of caregiving … that scale

The graphic above illustrates why enabling a caregiver to configure a PVAI to serve their needs and someone else’s constitutes a great good—the numbers.

Whereas today’s self-service model consists of one user interacting with one host-controlled app, every interaction between a care recipient and a PVAI serves at least two users at a time.

As adoption grows and trust builds, user-controlled services (UCS) will naturally evolve and expand. They’ll scale from a 1-to-1 caregiver-to-care recipient model to a many-to-1 model (e.g., several siblings supporting one parent), to a 1-to-many model (e.g., one adult child or paid caregiver supporting multiple aging relatives or clients), and soon to a many-to-many model (e.g., an in-home care organization with thousands of caregivers using PVAIs to monitor and support thousands of care recipients).

Tech-savvy readers will recognize how “network effects” will drive scaling … and scaling is needed. A recent AARP report states that 59 million Americans annually monitor and care for relatives aged 18 and over, most of whom are over 50. The same report finds that 24% of these caregivers care for more than one person, meaning the number of care recipients potentially benefiting from PVAIs could reach ~73 million.

Together, these numbers suggest that PVAIs could benefit 132 million personal caregivers and the loved ones for whom they care!1

To this, we should add professional caregivers employed by organizations like Home Instead, a provider of ADL (activities of daily living) support sans medical care, and BrightStar, a provider of ADL support and medical care. Claude.AI estimates that approximately 2.3 million professionals currently provide in-home care to roughly 4.5 million Americans each year.2

Add these professionals and their clients to the total above, and the potential number of Americans who could benefit from PVAIs with humans-at-the-helm grows to over 139 million—more than 50% of the U.S. population aged 18 and older!

And that’s only in the United States. Claude.AI estimates that 2.4 billion people outside the U.S. have “meaningful internet access in developed economic conditions”3 which—using ratios comparable to those in the U.S.—would represent another 953 million potential beneficiaries of this virtual architecture for e-care!

Finally, we should expect the PVAI + UCS platform to evolve to benefit other users (e.g., younger care recipients) and support other apps (e.g., voice-first control of a user’s online experience) just as the internet grew, soon after its launch, from online publishing and e-commerce to support countless other apps and users.

The Mission Multiplier effect

Numbers alone, of course, aren’t the whole story. At a time when many believe social media and other mobile apps are doing more harm than good, it’s critical to call out the exemplary mission of caregiving.

A caregiver’s core mission is to help others. With PVAIs + UCS, caregivers can finally leverage the power of AI + the internet to help others as well as themselves.

In our view, new “apps for good”—caregiving—will multiply the beneficial impact of scaling. Just as few imagined the vast mix of apps the web would support, an enhanced web supporting both caregiver and care recipient users will inevitably lead to an explosion of e-care apps benefiting millions of new users.

We predict, for example, that configurable PVAIs will help family and professional caregivers support older adults and others in their struggles with:

• Scheduling and remembering daily tasks, meds, and appointments,

• The risks and impacts of social isolation and loneliness,

• Preventing, detecting, and responding to life-threatening falls,

• Monitoring and managing myriad chronic conditions and diseases, and

• Collecting actionable data to improve the timeliness and efficacy of care.

This list of potential benefits is obviously high-level and must be supported by data. That said, the same critique could have been made—and was—when experts first predicted the web would benefit billions by creating a global “e-commerce” market 30+ years ago… which clearly happened.

Future posts will describe why we are so confident that this new e-care market will soon scale in similar fashion!

Yesterday, we described a new model for e-care that extends the legacy, self-service web to support configurable voice-AI agents. Here’s the graphic:

Quick recap:

This model introduces two key innovations: Proactive Voice-AI Agents (PVAIs) that initiate conversations with care recipients on behalf of their caregivers throughout the day, and User-Controlled Services (UCS) that let caregivers provide the agency and intelligence to configure those dialogs for both their needs and their loved one’s needs.

Together, these innovations will transform today’s reactive-only, host-controlled model into a new one—where caregivers are “at-the-helm.”

The power of proactive, user-controlled VAIs

As we described in Post 2, today’s mobile apps and services are all designed and controlled by a service or app owner. While we can choose when and how to visit a website or open a mobile app, every other user experiences the same interface and, almost always, the same content.

This new e-care architecture changes everything.

An ambient, proactive VAI means the care recipient no longer needs to know how to use a smartphone, check texts, or even “wake up” a VAI with a command.

A proactive VAI—configured by someone else—will be the first truly effortless UI for the millions of users with disabilities who currently cannot access or benefit from the online services that able-bodied users take for granted. No typing or tapping, no struggling to read a small screen or remember commands—just natural conversations initiated by a VAI that’s been configured by their caregiver.

It’s as if someone were in the room with them—speaking with the agency of their caregiver—talking naturally throughout the day.

For care recipients, it finally tears down the persistent barriers of the “digital divide.”

For family caregivers, a proactive VAI is the first digital connection they can truly trust … because they control it.

Combined with smart-home sensors and apps, it will give caregivers 24/7 agency and peace of mind when they can’t be there in person or can’t afford to pay someone else to be there instead!

Subscribe now

Giving users control of their online world

The idea of users having some control over their online experience isn’t entirely new. It began with user-generated content (UGC).

UGC allows users to upload content they’ve created—images, videos, audio, text—and interact with others based on that content. It spawned some of the most popular, profitable apps billions of us use daily: Facebook, LinkedIn, Airbnb, YouTube, and countless dating apps.

The UCS model goes much further than UGC. It enables a user to control not just the content another user sees or hears, but to define the app or service itself—what a PVAI talks about and when—and to control the invaluable data it generates and collects.

UCS will enable family caregivers to configure a PVAI to serve both as their proactive “voice” and as the voice of other care-providing services—human or automated—such as:

• A physician or service informing their aging parent of test results,

• A smart-home device announcing that someone’s at the door, and

• A motion detector proactively checking if an older relative is “okay” because they haven’t gotten out of bed yet.

In future posts, we’ll explore each of these potential additions to the PVAI+UCS solution—professional caregivers, smart-home devices/apps, and passive in-home sensors—in detail.

You’ll see that PVAIs can and should augment all of these point solutions, but only if and when the family caregiver remains at-the-helm!

Subscribe now

We hope you’ll subscribe and join us on this journey...

What a caregiver-configured VAI could do

A VAI set up by a caregiver to support a care recipient would help the caregiver by summarizing what the care recipient said and did during the day.

It would alert the caregiver—via the channel of their choice—whenever the care recipient says something (e.g., “I need help with my new meds!”) or does something (e.g., “I’ve fallen and can’t reach my phone!”) that requires immediate or in-person assistance.

An enhanced, proactive VAI could monitor a loved one virtually around-the-clock and collect data on behalf of a caregiver. The potential types and uses of such data are expanding daily.

Certain data, for example, could trigger critical alerts, while other data would support the creation and modification of care plans tailored to each care recipient’s unique needs—all easily configurable by the caregiving team.

A VAI that speaks first

Imagine configuring a remote VAI to proactively talk with your loved one about things you think matter, at times you think are best. For example, your VAI might say:

The first example might prevent your mother, who lives alone, from falling—again. The second might delight her because dementia means she wouldn’t have remembered Jenny’s birthday otherwise—and she might have struggled to schedule the call even if she did!

The graphic below illustrates the variety of check-ins, reminders, streaming content, and smart-home tasks that a caregiver could configure for a proactive VAI to guide and comfort a care recipient throughout each day.

A voice-AI agent configured this way would provide companionship and structure to the daily life of a remote relative or friend while giving them a feeling of connection and support from their loved ones.

The missing piece

Despite recent innovations like Small Language Models and Action Models (SLMs and SAMs) that enable friendly voice chatbots to talk with users about a variety of topics, no such proactive VAI exists today—one designed to serve aging or chronically ill persons that can be configured by a caregiver or caregiving team.

Reimagining the online model for the greater good—caregiving

Given the internet’s legacy self-service model, today’s voice chatbots only react to and engage with us when we open a mobile app or visit a website. Put simply, today’s apps can’t initiate an interaction with another end-user, regardless of who it is.

Our proposed solution does exactly that:

While it may surprise some, this capability is not only possible but straightforward with today’s technology.

The next graphic shows how this new e-care model extends the legacy e-consumption model to support configurable, proactive VAIs for caregivers.

This new model incorporates two key extensions of the legacy web model.

First: Proactive Voice-AI Agents (PVAIs)

Care recipients, shown on the left, can communicate effortlessly and exclusively via voice largely because the VAI no longer waits for them to “wake it up” before engaging.

Instead, it proactively starts dialogs with the care recipient—on behalf of their caregiver—throughout the day. We call this new interface a “proactive voice-AI agent” or “PVAI” to distinguish it from today’s reactive-only VAIs.

Second: User-Controlled Services (UCS)

An individual caregiver or caregiving team, shown on the right side of the graphic, can provide the agency and the intelligence that the PVAI lacks in the form of scheduled dialogs tailored to both their own needs and those of the care recipient.

We call these powerful new capabilities “user-controlled services” or “UCS” to contrast them with the only online model we’ve ever known: host-controlled services (HCS).

Subscribe now

Today, we introduced two innovations—PVAIs and UCS—that finally make the web work for caregivers and care recipients alike. Tomorrow, we’ll explore how these innovations work in more depth and show how other stakeholders and technologies can supplement and support the family circle of care.

We appreciate the power of this consumer-centric model because we enjoy the freedom and convenience it provides. We can start and end every online session using whatever device and interface we have at our disposal, whenever and wherever we choose.

The pervasive self-service model of smartphones and touchscreens fits the e-commerce model and other digital markets—including social media and self-care—perfectly.

The graphic below shows how this internet + AI model has delivered once-unimaginable value and convenience to consumers of all stripes since the mid-1990s.

Self-service, self-care versus caring for someone else

When it comes to e-caregiving—what we call “e-care”—there’s a problem.

Self-care apps fit today’s self-service consumer model because they enable users to care for themselves, not others. Caring for someone else, however, is the antithesis of caring for oneself.

Self-care is also a singular, transactional customer experience, whereas caregiving is fundamentally a shared activity. Caregivers proactively “give” or share their time and energy to care for another person—a relative, friend, or client.

In short, caregiving—caring for someone else—is the virtual opposite of self-care.

Caregiving done right is holistic and based on a relationship developed between two people over time.

Whereas self-care using the web—or “care-consuming”—is based on repeated transactions between an individual end-user and an app that supports prompt-response interactions, usually via text, about a disease or condition.

Once a self-care session ends, so does the care-receiving—until the user activates the app again.

Today’s caregiving, by contrast, still happens in real-time, in-person or, when caregivers can’t be physically present, via a phone or video call.

And when they can’t connect in-person or via phone, family caregivers don’t stop thinking about their aging relative. Instead, they worry whether their loved one is “ok” and consider whether there are other ways or other people who might help watch over and support their care recipient in their absence.

What if a next-gen chatbot could proactively talk with and support someone else on behalf of a caregiver or caregiving team?

The untapped potential of a virtual, configurable presence

For a caregiver to monitor and support an aging loved one who’s two blocks or two states away via a voice-AI agent (VAI), the path is clear:

Set up and configure a VAI to “say” what the caregiver would say if they were physically present, on days and at times that work for both the caregiver and the care recipient.

By now, caregivers should be able to configure VAIs to proactively do this and much more, including:

• Converse with a care recipient about a variety of topics

• Check in regularly about how they’re feeling and other timely concerns

• Help them access and navigate online services (e.g., listen to music) or control smart-home devices (e.g., TVs or door locks) via voice

• Remind them to do things that might please or benefit them (e.g., go for a walk, call a friend, take their meds, or stay hydrated)

A caregiver, in other words, should be able to set up and configure a VAI not just for themselves—as we do today—but to serve the needs and wishes of their older parent or, if they’re a paid professional, an aging or chronically ill client.

Tomorrow: The two innovations that will finally make the web work for caregivers and care recipients alike. Stay tuned …

Nonetheless, healthcare tech startups have yet to provide much other than (a) tele-health solutions for virtual visits and (b) point solutions for self-care of specific diseases or conditions – neither of which addresses the myriad needs of tens of millions of family caregivers and their loved ones who are aging-in-place.

Instead, the global crisis in caring for people as they suffer from disabling chronic conditions (e.g., dementia), debilitating falls and the social determinants of health (e.g., isolation and loneliness) continues to grow.

This crisis is particularly acute in the United States where it is substantially more expensive yet arguably less efficacious – as measured in human outcomes – to care for someone and to receive professional care than elsewhere.

Most readers will have experienced this crisis firsthand before they turn 50 as caregivers or care recipients or both.

By then, most of us have confronted the challenges of monitoring and supporting an aging relative or friend from a distance and, in the process, learned that we have few, if any, affordable options other than to add “personal caregiver” to our other roles as parents, partners, friends and employees.

Relying instead on paid, professional caregivers is expensive and made even more problematic by the shrinking supply of qualified caregivers … which causes costs to rise in a never-ending spiral.

Even those lucky enough to have a paid or volunteer “workforce” to help them care for a loved one are often consumed by the time and effort required to manage their “caregiving team”!

Everyone’s talking about it … everywhere

If you’ve yet to serve as personal caregiver or care recipient, you’ve likely read or heard about it in the constant drumbeat of anecdotes and analyses coming through on all types of media.

In his widely acclaimed book, Being Mortal: Medicine and What Matters in the End, for example, Dr. Atul Gawande describes the intractable challenges facing Americans as they age and their caregivers.

Gawande concludes that biomedical innovations have enabled us to lengthen life but have yet to provide digital solutions that enable care recipients to maintain their independence and quality-of-life as they age or to help caregivers monitor and support them … or both.

Along the way, he acknowledges the obvious – that any such solution that works will launch a massive, new market:

If scientists came up with a device—call it an “automatic defrailer”—that wouldn’t extend your life but would slash the likelihood you’d end up in a nursing home or miserable with depression, we’d be clamoring for it. … Medical students would be jockeying to become defrailulation specialists, and Wall Street would be bidding up company stock prices.

While his imagined automatic defrailer may never come to pass, it’s well past time for AI + the internet to ease the burdens of aging for all – both care recipients and their family caregivers.

The point of departure …

This all begs an important question: Why hasn’t the past 30 years of investment and innovation in the digital-virtual space resulted in an array of mobile apps and online solutions that help caregivers monitor and support their aging relatives or clients at home while simultaneously serving those for whom they care?

It’s not for lack of technology or vision. Large language models now support conversational chatbots tailored to specific users and tasks.

New voice chatbots are providing mental health counseling that is less expensive yet, for some, more convenient and effective than in-person therapy (e.g., BetterHelp). A few of these chatbots are even being added to robots that seem to see and move slowly!

Despite these and other advances, however, leading voice-AI agents (e.g., Alexa and Siri) have not evolved to help individual caregivers monitor and care for someone else—on the caregiver’s behalf—via a friendly yet automated voice.

What to expect …

In this Substack, John and I will explain and explore why the virtual world has yet to find its “voice” … and how it can and will find it soon!

We’ll identify the foundational barriers in the current AI+internet model that have delayed the development of an “e-care” market for connected, voice-AI agents that work for and with human caregivers and their care recipients.

Then, we’ll outline our vision of what’s possible now. We’ll describe how two tech innovations will enable entrepreneurs and care providers to overcome these barriers and launch groundbreaking apps and business models that will benefit millions, eventually billions, of caregivers and care recipients alike.

We hope you’ll read on!