is increasingly Common experience: You walk into the kitchen, muttering under your breath, when you hear a disembodied female voice say, “Sorry, I misunderstood that.”

We can all agree that Alexa’s tendency to eavesdrop is a bit creepy at times. But, is it possible to take advantage of that capacity to improve our health? That is the question that researcher David Simon and his co-authors tried to answer in a recent article published in Cellular press.

Simon, a legal ethicist at Harvard University, and his team envisioned a hypothetical near-future scenario in which Alexa came equipped with the power to diagnose cognitive conditions like Alzheimer’s and dementia simply by analyzing a person’s speech patterns. elderly.

In doing so, they tried to imagine all the benefits such technology could offer and all the ethical problems it could create.

“These technologies are already beginning to exist,” says Simon. Reverse. “It’s a rich area for future research.”

Calling Dr. Alexa

When diagnosing conditions like Alzheimer’s, time is of the essence. Symptoms of the condition, which include memory loss, confusion, and personality changes, often come on gradually, making early diagnosis an incredibly difficult task.

“[Technologies like this are] coming faster than the law is equipped to deal with.”

Most people with Alzheimer’s live four to eight years after diagnosis, but some patients whose symptoms are caught early can live as long as twenty. The ability to catch symptoms before they escalate could give the person and their family time to “address the problem while they still have the ability to do so,” says Simon.

But cognitive conditions can be particularly difficult for doctors to diagnose the old-fashioned way. A typical checkup takes less than an hour. During that time, a doctor could easily miss some of the more subtle symptoms of early Alzheimer’s disease. Similarly, family, friends, and even the person themselves may miss early warning signs, especially if they are not constantly in contact with each other.

Personal digital assistants, on the other hand, are always present and listening. With the right software, these devices could analyze a person’s speech data over many years and detect the warning signs of cognitive decline, such as forgetting words or losing fluency, before anyone else. This data could also help a doctor confirm a patient’s diagnosis.

So does this mean doctors have to hand over the prescription pad to Alexa? Not so fast.

The risky ethics of digital eavesdropping

There are some obvious problems with receiving medical advice from your personal digital assistant. The first is the possibility of misdiagnosis.

“Innovators like to innovate. They are motivated to make a product that is marketable.”

The AI ​​certainly isn’t perfect (except for DALL-E mini, of course). A program designed to track cognitive decline in the elderly may be confused if it encounters the speech of an adolescent or young adult, and a program trained in neurotypical speech patterns may have trouble accurately assessing a neurodivergent person.

Then there is the question of privacy and consent. A listening device in your home will likely pick up some juicy gossip along with helpful diagnostic data. Patients may not want sensitive information about their love life, for example, to be presented as evidence of cognitive impairment to their doctor.

And the slope becomes even more slippery around progressive cognitive diseases. A person with Alzheimer’s may not realize that they are being monitored by an omniscient AI, even if they were the one who suggested using the program in the first place.

But topics like these sometimes don’t even enter the conversation on the development side.

“Innovators like to innovate,” says Simon.

“They are motivated to make a product that is marketable.”

It is up to ethicists, policymakers, and medical regulators to deal with these potential drawbacks. And Simon thinks they should start fighting as soon as possible.

On the horizon…

The idea that Amazon (or a similar company) uses its digital assistant as a medical tool is not far-fetched. These devices are already equipped to listen for voice commands and tuned to receive environmental data from around the house; calibrating them for diagnostic purposes could be as simple as a software update. And getting such technology approved wouldn’t be too difficult under current US guidelines.

From a regulatory perspective, it’s much easier to get FDA approval for so-called “medical devices,” which include everything from glucose meters to Fitbits, rather than drugs.

Additionally, Amazon’s recent purchase of One Medical positions it to seriously enter the healthcare industry. The megacompany already acquired online pharmacy PillPack in 2018 and made its first foray into health care with its Haven pilot program (which closed last year).

Because of this, Simon believes now is the time to start thinking about the implications of letting Alexa diagnose his grandmother.

“Technologies like this are coming. And I think they are coming in faster than the law is equipped to fully address,” she says.