For the first time in six years, our hearing loss community came together online for a national Hearing Loss Association of America (HLAA) virtual meeting. It was such a joy to see so many familiar faces and new ones from all over the country. Alongside Roxana Rotundo and Holly Cohen, I had the pleasure of co-hosting an evening built around one of the most talked-about topics of the moment: artificial intelligence and what it means for those of us living with hearing loss. We then shifted gears for community sharing about the unexpected gifts of living with hearing loss.
Below, I share highlights from the meeting, but you can also watch the replay here. The full Research Symposium is available to watch here.
Understanding the Risks and Rewards of AI in Hearing Care
The meeting began with a recap of the research symposium from this year’s HLAA Convention in Louisville, “The AI Revolution in Hearing Health: Smarter and Greater Access.” I chatted with Dr. Jan Blustein, who moderated the symposium. She is a physician, health services researcher, longtime HLAA member, and a person with hearing loss herself.
We discussed the highlights from each of the event’s three speakers:
- Dr. Nicholas Lesica, a bioengineer from University College London
- Cole Morris, co-founder of the hearing aid company Fortell
- Dr. Samantha Robler, an audiologist and public health researcher.
As a reminder, HLAA does not endorse any products or brands discussed during the meeting. Dr. Blustein disclosed that she consults for Fortell, Inc.
What Is AI, Really?
We started with the basics. What is AI? If your mind jumps to robots or ChatGPT when you hear “AI,” you’re not alone — but that’s not what we’re talking about here. As Jan explained it, the AI showing up in hearing aids is simply a tool for detecting patterns in data. Imagine you want to identify which children in a group have ear infections.
For example, an audiologist could examine each child one by one. Or you could train a computer on tens of thousands of prior cases — otoscope images, temperatures, known outcomes — until it learns to recognize the patterns that signal an infection. That ability to learn from enormous amounts of data is the “secret sauce” of AI. It isn’t magic, and it isn’t foolproof, but it’s powerful.
Can AI Help Hearing Aids Better Handle Noise?
Understanding speech in noisy places is the number one challenge most of us with hearing loss face. At the symposium, Cole Morris made a provocative point, saying that simply having “AI” in a hearing aid tells you almost nothing, because AI can be just about anything.
In hearing aids, what actually matters is whether the technology improves the signal-to-noise ratio — making the speech louder without making the background louder. Jan explained that people with typical hearing do this automatically; the brain suppresses the barking dog and the clattering silverware. For people with hearing loss, that filtering doesn’t always happen on its own. Standard hearing aid programs, even those with AI, aren’t always much help.
The good news is that some newer AI approaches are having more success. They split incoming sound into two tracks — a speech track and a noise track — and then remix them, like a DJ adjusting two faders: pushing the speech up and pulling the noise down. This is an exciting avenue to watch, but challenges remain. Fitting all this computing power into something the size of a fingernail that runs all day is quite a challenge. Plus, what is noise to one person may not be noise to another. So far, hearing aids have not been able to solve what Jan called the “intent” problem.
The Intent Problem
The intent problem is exactly what it sounds like. Our devices can’t yet read our minds. Your hearing aid may be able to separate speech from noise, but how does it know which voice you want to hear? If a loud talker sits down at the next table in a restaurant, the technology may happily amplify them too. Noise can be subjective as well. Is that background music adding to your experience or an annoying distraction? The answer may be different for each of us or in different situations.
Today’s answer is for devices to use directional microphones that favor whoever is in front of you. This is great for a one-on-one “date night,” as long as you’re willing to turn your head. Researchers are working toward something more remarkable: the ability to pick out a specific voice, like your spouse’s, from the crowd. That’s always been my dream: the ability to train our devices to pick out the sounds and voices that matter most.
We’re not there yet, but it’s coming.
AI and Access to Care
Dr. Samantha Robler’s work came at the topic from a completely different and deeply moving angle: using AI to reach people who can’t easily get to an audiologist. In many rural and low-income areas, there simply aren’t enough specialists, and treatable ear problems in children go undiagnosed.
The stakes are sobering. According to Robler, about 75% of childhood hearing loss in rural and underserved communities is preventable, often stemming from untreated chronic ear infections. Yet most affected children are never identified. She noted that untreated hearing loss makes kids three times more likely to repeat a grade and less likely to graduate.
Robler’s team is building a low-cost tympanometer that uses AI to screen ears as normal or possibly infected — no audiologist required — so a school nurse can flag children who need further care. Her research already spans roughly 100 schools and 20,000 children in rural Appalachia and Alaska. Robler’s vision is for AI and human specialists to work together, with AI handling the clear-cut cases quickly and people handling the complicated ones. Her motto says it beautifully: “Right care, right place, right time.”
Don’t Be Swayed by the “AI-Powered” Label
Every hearing aid ad these days seems to promise something “AI-powered,” so I asked Jan how consumers should evaluate those claims. Her answer was one of the most useful takeaways of the night: since almost everything uses AI now, the real question is what the AI is actually doing. Is it working to solve the speech-in-noise problem? And if so, which devices do it best?
That turns out to be a hard question to answer. Head-to-head scientific comparisons between devices are few and far between, and manufacturers haven’t been eager to produce them. For a long time, as Jan put it, “whoever shouts loudest wins.” That’s finally beginning to change, and a key message from the symposium was that consumers should demand evidence, not advertising. For now, the bottom line is that you often need to try devices to find out what works for you.
One helpful, impartial resource is HearAdvisor, run by the team at Hearing Tracker. They independently test a wide range of hearing devices, including over-the-counter models, and publish the results at HearAdvisor.com.
What About the Risks?
Jan discussed a few concerns worth keeping in mind. Our hearing devices contain microphones and could, in theory, be hacked. They store personal medical data that could potentially be shared in ways we wouldn’t want. And AI makes mistakes, which means “AI-enabled listening” can simply be wrong.
The takeaway isn’t to reject the technology, but to stay informed and ask questions about how AI is working in your devices, what data they collect, and which settings you can and cannot control. And, as Jan pointed out, some of these concerns aren’t unique to AI — they’re questions we should be asking about connected hearing devices generally.
We Have a Voice in What Comes Next
Dr. Lesica closed his symposium talk with a call to advocacy, so I asked Jan what people with hearing loss — and organizations like HLAA — can do to help shape where this technology goes. Her answer was very much in the patient-centered spirit I love about our community.
We can demand evidence over advertising and ask manufacturers for solid, real-world results. We can tell researchers what we want, in venues like the Research Symposium, so they know what to build toward. We can encourage our audiologists to get involved in research and to speak up for what patients need when they meet with manufacturers. And we can tell anyone who will listen that we want science and facts, not just marketing claims.
We don’t have to be passive consumers of new technology. We can help shape the questions researchers ask, the evidence manufacturers produce, and the technology they ultimately build. As Jan summed it up: when it comes to speech in noise, help is on the way.
A big thank you to Jan, for skillfully summarizing the symposium on the call. It was a terrific discussion!
The Silver Linings
The evening closed by turning from technology to community.
In our “Silver Lining Sharing” segment, attendees reflected on the unexpected gifts of living with hearing loss. People spoke about the strength and warmth of the hearing loss community itself, the deep empathy and care they’ve developed for others, and the many tools — meditation among them — that help them live well.
It was the perfect note to end on. AI may open exciting new possibilities for hearing technology. But technology is only part of the story. What matters most is how we use it to stay connected, participate fully, and make sure our voices are heard.
Replay Available
Missed the meeting or want to listen again? A recording of the virtual meeting is available here. The full Research Symposium is available to watch here.
Readers, what do you hope AI will — and won’t — do for people living with hearing loss?
Related Posts
Connect with us on Facebook and Twitter!
I wish you would address cochlear implants . I have two and still have a lot of problems hearing. I feel I have done everything there is to hear better but still can’t hear in groups.
Thank you for your suggestion. It is an important one.