This article is part of the On Tech newsletter. You can sign up here to receive it weekdays.
Are voice recognition technologies like Alexa helpful in medicine or are they hogwash? For now, the short answer is a little of both.
Microsoft as well as other tech companies like Google and Amazon have big ambitions to transform the industry with artificial intelligence technologies, including in voice recognition programs and efforts to identify signs of illness and disease.
The big hope of technology in medicine is that it can help make us healthier and improve America’s expensive and often ineffective and unjust health care system. The message that I have heard from medical experts is that there’s potential there, but there is also a lot of hot air.
The hope of medical Alexas:
For years, doctors have used Nuance’s transcription software to speak notes about patients and convert them into text for medical records. In theory, that frees doctors from having to do paperwork so they can spend more time treating us.
Nuance and other tech and health care providers want to do much more with our voices. One idea is that microphones might record (with permission) interactions between physicians and patients and log the relevant details into medical files — without much human involvement. Computers would also be smart enough to order any necessary tests and handle billing.
This sounds cool and perhaps a little creepy. These ideas are still under development, and it’s not clear how well these medical Alexas would work. But Dr. Eric J. Topol, a professor of molecular medicine at Scripps Research and the author of several books on technology in medicine, told me that voice recognition systems are one of the most consequential uses of artificial intelligence in health care, at least in the short term.
At Cedars-Sinai, a health system in Southern California, most hospital rooms have been outfitted with voice activated devices, said Darren Dworkin, the organization’s chief information officer. For now, the devices are mostly used for relatively mundane interactions, such as a nurse asking a device to show a patient a video on preventing dangerous falls.
Dworkin said that he was most optimistic about using voice and other technologies to automate administrative work, such as authorizing insurance for medical treatments and sending tailored text messages to patients.
Dworkin said that those uses of technology might not be what many considered a wow factor, but that busywork was a huge cost and challenge in health care.
“Not everything has to be state of the art,” Dworkin said. “Don’t let the simple stuff pass you by.” (Another vote for the importance of boring technology!)
Where hope meets harsh reality:
Just about every technology used in health care — and many other fields — promises to reduce administrative work and costs. And yet, health care expenses and bureaucracy in the United States mostly continue to go up.
Dr. Dhruv Khullar, a physician and assistant professor of health policy and economics at Weill Cornell Medicine, said that he was optimistic that voice tech and artificial intelligence could reduce administrative burdens and help patients. But he said that his hope was not yet backed by rigorous proof.
“There is not a lot of evidence at this point that A.I. reduces costs or improves health outcomes,” Dr. Khullar told me. (I borrowed the “medical Alexas” line from him.)
I asked these health experts an overarching question: What role should technology play in tackling the root problems of American health care?
They largely agreed that advances in technology could help reduce costs and improve the quality of service in our health care system, but that it was not a silver bullet for our biggest problems.
“I would say, it’s part of the answer but not a large part of it,” Dr. Khullar said.
(And read more from DealBook: How has Microsoft mostly avoided the government’s antitrust attention? My answer: Microsoft’s essential technology is mostly dull. That is a good thing.)
Hacking technology, with long distance operators
Last week, I pointed to a terrific article about Indians adapting to expensive mobile phone calls by coming up with new ways to communicate that involved hanging up mid-ring. An On Tech reader, Morris Fried of Somerset, N.J., wrote to us about his family’s missed call communications system from decades ago:
Your note about using missed calls for communications in India stirred old memories of the same technique in this country. (I will be 75 next month.)
When I was a child, we would drive back home to Philadelphia after visiting my grandmother in Brooklyn. My mother would then call the operator and request a person-to-person long distance call to her own name at my grandmother’s phone number.
My grandmother would answer the phone and tell the operator that my mother was not there. My mother thereby succeeded in informing her mother that we had arrived home safely without incurring the then not-insignificant expense to us of a long distance telephone call.
Before we go …
Hugs to this
“If you’ve always wanted your own haunted Victorian child in the body of a small dog that hates men and children …” I laugh-cried at this extremely detailed description of Prancer on Facebook and his MANY peculiar habits, posted by a New Jersey pet adoption league.