Conversational and generative AI

Asking ChatGPT a question on an iPhone returns a detailed answer.  Of course, it’s slightly different when asked again. The question: “What devices are useful for monitoring older adults in their home?”  The­­ categories (and sub-categories) were not surprising – you can give it a try yourself.  They included medical alerts, smart home devices, cameras and video monitoring, remote health monitoring fall detection sensors, GPS tracking devices, medication management, environmental monitoring.   On the iPhone, adding companion robots – and an observation: “These devices, especially when used together, can create a safer and more supportive environment for older adults living independently.”

An old report, the core concept of Connected Living was excellent and predictive.  Thirteen years ago, AARP sponsored research that posed questions about technology’s future role in connecting older adults with families, resources and each other. With input from 30 industry experts, the research attempted to determine how technology could better serve older adults moving forward. The result was a 2011 report called Connected Living for Social Aging: Designing Technology for All. You won’t find it on AARP’s website – it’s too old.  But it is very interesting, especially given that year's low technology adoption and extremely limited use among older adults compared to today. The report accurately predicted the major role technology would take in their lives as they aged, though experts were not exactly sure how.

The month of May -- and the hostility about AI overflowed.  Given the pace of change in AI technology – both the software and its rate of adoption – it’s curious that recently the Wall Street Journal published an aging survey about what customers don’t use and/or like about chatbots. These observations include the usual: ‘hallucinated’ answers; lack of customer awareness that they are talking to a chatbot (really???); the chatbot is too nosy. Or it asked too many questions; or couldn’t handle two questions. Which would make this article, like much of media coverage of AI, sound negative. Too late, adoption happened anyway. This is a commentary, perhaps, on the nature of news media in general, who either are mirroring the AI skepticism in the public, or more typically promoting it. But clearly with chatbot adoption, the public is paying new attention. Sigh.  Here are the four blog posts from May, 2024:

It’s 2024 -- chatbots, yuck? Given the pace of change in AI technology – both the software and its rate of adoption – it’s curious that recently the Wall Street Journal published an aging survey about what customers don’t use and/or like about chatbots. These observations include the usual: ‘hallucinated’ answers; lack of customer awareness that they are talking to a chatbot (really???); too nosy. Or it asked too many questions; couldn’t handle two questions. Which would make this article, like much of media coverage of AI, sound negative. Too late, adoption happened anyway. This is a commentary, perhaps, on the nature of news media in general, who either are mirroring the AI skepticism in the public, or promoting it. But clearly with chatbot adoption, the public is paying new attention. 

The report is published, the feedback positive, observations strike a chord.  Necessity will drive AI usage in care work across all five care types (healthcare, home health care, home care, senior living, and Skilled Nursing Facilities). Issues of worker shortage, staff burnout, or migration of care work into the home will result in broader deployment of AI technology (whether explicitly or inside other software tools). And regulatory initiatives will help overcome trust issues for consumers. Over the next few years, care organizations will make more disciplined use of their own data that an AI technology such as a chatbot can access or present to a caregiver. The changes that are most likely within the next five years? See today-future comparison chart below and check out the report here.