Using AI to help with important life decisions

East Kildonan

Hey there, time traveller!
This article was published 21/05/2025 (280 days ago), so information in it may no longer be current.

Over the past 18 months or so, there has been an explosion in news about artificial intelligence (AI). It can be confusing, and perhaps a little scary, if you’ve seen a few science fiction movies.

As a financial writer, I receive story ideas about AI every day. Many are trying to capitalize on the craze, but don’t use AI for much, if at all.

However, sometimes I get to talk to someone who is using AI to help people with important life decisions. Waterlily co-founder and CEO Lily Vittayarukskul is one such person. A NASA intern in her early teens, Vittayarukskul had an unpleasant experience that many of us are ill-prepared for — caring for a sick relative.

Vittayarukskul’s aunt was suddenly diagnosed with terminal cancer. That forced her family to have many difficult discussions about her care, at a time filled with grief and stress. A decade later, some relationships are still frayed.

Vittayarukskul channelled that experience into Waterlily, which helps financial advisors, insurance companies, and other professionals have easier conversations with clients about saving for their long-term care needs.

Waterlily uses artificial intelligence to analyze more than 500 million data points, including the experiences of thousands of families who have previously navigated the process. The system compares people with those most like them. While it is currently only available in the United States, Vittayarukskul hopes to bring it to other countries.

Because envisioning declining health can be a difficult and stressful process, people avoid thinking about it. That means they don’t plan for it; a dangerous strategy when today’s long-term care costs could more than triple by the time some adults need to pay for it.

So, we have to strategically save for it. Long-term savings discussions are often held with financial advisors, so Waterlily focused the design on them. They prioritized making those discussions as straightforward as possible.

“We found that it was a particularly good use case for predictive AI,” Vittayarukskul said. “How do you make it relevant to the client sitting in front of you? How do you gather all the right data sets so that you can have an accurate view as best as you can?

The client answers basic health and family history questions for three minutes or so. Then they discuss their hopes whether it be staying in their home for as long as possible or moving into assisted living at a certain age. Even the amount of available family support is factored in.

The AI analyzes the data, which includes the cost of local assisted living and long-term care options, and produces a simple, yet personalized, plan. If the initial result is unacceptable, the advisor and client can change the inputs to see if something more realistic results.

With the numbers there in black and white, the discussion becomes about planning for those long-term needs. It takes some of the stress out of what can be a difficult process.

“What we’re trying to do is save these families (perhaps) hundreds of thousands of dollars on what this event is going to cost,” Vittayarukskul said. “It serves the overall economy around how we sustainably age as a population, and how do we control that cost sustainably for families that either pay out of pocket or the insurance industry that’s trying to figure it out?”