10
Loading video...
By Danielle Sutherland
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
A retired teacher may still hear from former students who want help with a lesson.
Imagine if those students could search a collection of the teacher’s recorded lessons by asking a question in their own words. Artificial intelligence (AI) could help them find the part of a lesson they need, and they could follow up with the teacher if they were still stuck.
I believe this is one of AI’s most promising uses. It can give people more ways to share what they know, speak with confidence, and stay involved in the parts of life they value.
Read More on Leadership Insights
To see whether it delivers on that promise, developers should test it with the people they expect to use it, make clear when it can’t answer a question confidently, and measure what people are able to do with its help.
Consider what people want to keep doing as they age. In a national survey, AARP Research found that 75 percent of U.S. adults aged 50 and older wanted to stay in their current homes for as long as possible.
For someone who wants to stay in their neighborhood, being able to keep track of a weekly activity or arrange a ride to see a friend can matter.
Don't Lose Sight of AI's Goals and Capabilities
If a tool is meant to help someone stay connected, a good test is whether it helps them get to the places and see the people they want to.
Businesses can lose sight of those goals when they measure AI’s success mainly by how much work it takes over. A company that moves customer requests to a chatbot may see fewer calls. But a customer who gives up halfway through an unfamiliar form disappears from that count.
I would judge the change by how many people completed the form and whether those who needed help could reach someone. Both measures tell the company something about the service its customers received.

...
The teacher’s recordings offer a simple example of what it means to make uncertainty visible. If a student asks how to solve a math problem, the tool could point to the exact moment the teacher explains it.
If it can’t find an explanation that fits, it should say so and give the student a way to ask the teacher. The teacher decides what to share, and the student can see where the answer came from.
At work, the outcome might be whether someone feels ready to speak up. A person who hesitates to speak up in meetings might use an assistant to organize their notes, practice explaining an idea, or think through questions colleagues could ask.
Over time, the assistant might remember that the person prefers a short list of talking points, a preference they could review or change. A developer testing that tool could ask whether the person felt better prepared and whether they actually brought their idea to the meeting.
Choose a Goal Before Testing an Assistant
The same approach could help someone who sometimes loses track of plans keep up with a familiar routine. They might ask an assistant when their walking group next meets or use it to keep a list of questions for a medical appointment.
A family member could help set up reminders if invited, and the clinician would still guide decisions about care.
The person using the assistant would choose which activities they want help with and who can see the information they share.
I’d like to see developers ask participants to choose a goal before they start testing an assistant. After several weeks, ask whether they’re closer to that goal, whether they can correct mistakes, and how easily they can reach a person when they need help.
Businesses buying these tools should ask to see those results alongside estimates of time saved.
The result may be as ordinary as someone showing up at their walking group next Tuesday and picking up a conversation they’ve missed. If an assistant can help make that possible, it has done something worth measuring.
Jevity was built from two convictions. I saw that technology could help people stay independent, show up for the people they matter to, and keep doing what they love without replacing the human relationships that make those things worth doing.
David Jackson believed an assistant could grow with someone, remember what they care about, and become more helpful over time.
Together, we wanted to build something that would be judged by whether it helped someone with day-to-day goals or tasks, giving the person on the other end what they need: time to focus on what matters.
Danielle Sutherland and David Jackson work on Jevity SI, a conversational assistant designed to support people in their daily lives. All views expressed are the author's own.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines