The fear that machines will replace humans in the workplace is not a new one. In 1930, economist John Maynard Keynes conjectured that in the years to come, modern economies would face a new kind of affliction: what Keynes called “technological unemployment.” During the Industrial Revolution — one of the more notable examples of a radical paradigm shift in the labor market due to automation — textile laborers were supplanted by steam-powered machinery. Today, we no longer manually assemble cars in factories. The days of human telephone switchboard operators are long gone. And by some counts, roughly half of all low-skilled positions are expected to be replaced by artificial intelligence and automation.
It’s difficult to accurately gauge the threat of automation and AI to the human workforce. Some argue automation alone could spell the end for many service workers, such as cashiers, truck drivers, and retail salespeople. A recent study from the National Bureau of Economic Research — based on U.S. labor market statistics from 1990 to 2007 — revealed that for every robot introduced into the workplace, between 3 and 5.6 human workers lost their jobs, and the study predicted a deep decline in the employment-to-population ratio. With the added consideration of quickly growing artificial intelligence used to augment the machinery, it’s no wonder experts and laypeople alike fear for mankind’s ability to make a living in the future.
Yet there are equally many who contest this dire forecast. Some claim the statistics paint an inaccurate picture, adding that perhaps only 10 percent of jobs are at risk of automation and that, even then, technology also tends to create jobs, too. One contributor to MIT Technology Review argues this fear is merely hysteria generated by a variety of misconceptions, largely based on an overestimation of AI’s uses and abilities and on Hollywood notions of science fiction scenarios.
Still, the trepidation persists, and it’s not just limited to manual laborers and service workers. Many in the medical field fear that they may be under threat of losing their jobs to AI — and these concerns aren’t entirely unfounded. Personal health care robots, which could assist in managing personal medication schedules and tracking health progress, are predicted by some to replace domestic nurses. In addition, anesthesiologists may face upheaval, since Johnson & Johnson already has an FDA-approved device that automates the delivery of low levels of anesthesia.
On the whole, however, it’s not a matter of AI versus physicians, but one of AI plus physicians. Perhaps the first step towards realizing AI stands to do more good than harm in the health technology sector is deepening one’s understanding of what AI means pragmatically. A far cry from the hyper-intelligent entity often portrayed in science fiction, AI is in actuality only a tool used to assist and enhance physicians’ practices. Because of this, IBM suggests referring to it not as artificial intelligence, but as augmented intelligence, compounding the notion that AI is merely intended to enhance products and services, rather than replace the humans that offer them.
Analyzing patient data
One major issue facing physicians is how to use the enormous amounts of data to effectively treat patients. Electronic medical records (EMRs), sensors, and wearables collect huge amounts of information, allowing for both clinic- and home-based medical data gathering through continuous monitoring. AI can enhance doctors’ critical decisions by unlocking patterns from all of that new data. IBM Watson for Oncology synthesizes data from “over 290 medical journals, over 200 textbooks, and 12 million pages of text” to uncover methods of treatment that no single human could find on their own.
It’s simply impossible for humans to efficiently synthesize all of this information to generate useful insights, yet this is exactly the kind of predictive analytics that humans can use to augment their practice to prevent life-threatening complications and provide ultra-personalized treatments.
Eliminating mundane tasks
This uptick in data also means larger quantities of rote work for medical specialists. According to a recent report, radiologists, in particular, must review a new medical image every three to four seconds during a typical eight-hour workday; the time dedicated to viewing and categorizing these images might be better spent elsewhere. AI can help turn masses of information into actionable knowledge, which radiologists can then use to do what they’re trained to do. While an AI assistant is useful for sorting through huge amounts of imaging data, the final critical decision-making is still best left to the human specialist.
Additionally, administrative tasks and clerical work consume an unseemly amount of any physician’s workday. A recent study published in the Annals of Internal Medicine revealed that doctors now spend over two-thirds of their time doing paperwork — roughly double the amount they spent just a decade ago. AI can lessen the burden on doctors by handling repetitive, information-intensive rote tasks. Leveraging AI algorithms to organize doctor’s notes or chatbots to handle simple tasks like medication reminders and symptom assessments will ensure doctors can provide better diagnoses and treatments with less time and effort, thereby increasing overall health care efficiency.
All of this paperwork also reduces the amount of time physicians can devote to direct patient care. According to a Medscape report from last year, doctors spend an average of 13-16 minutes with each patient. By allowing AI to manage rote tasks such as research, data analysis, diagnostic assistance, and treatment options, doctors will be able to devote more time to their patients, completely redefining the doctor-patient relationship.
Medicine will always require a human touch
AI is already successfully integrated on a small scale. AI health assistants and chatbots use natural language processing and generation, as well as machine learning algorithms, to create a complex map of a patient’s condition, generate detailed symptom assessment reports, and set up live video consultations with general practitioners. It’s important to keep in mind that these technologies are intended to assist caregivers, not replace them.
Most importantly, medical practice and the art of healing will always need a human touch. Empathetic communication is proven to not only improve both doctor and patient satisfaction but even enhance diagnostic accuracy. Patients are more willing to provide a full history to physicians who engage with them on a deeper, more compassionate level, as opposed to those who provide prescriptions and other treatments alone.
In order to fully address the health problems patients face, health care delivery will need to find new ways to enhance the human-to-human relationship. Counter to conventional wisdom, AI can allow physicians to deliver more effective care and connect with their patients on a more human level by freeing them from burdensome administrative and analytical tasks.
The impact of AI on health care
The transformative impact of AI on the health care industry is not a potentiality — it is an eventuality. With any technological revolution, there are risks that would be foolish to ignore. But doctors and AI can work together without the loss of livelihood if we take the time to understand and prepare for the change that’s coming.
Doctors and patients alike stand to benefit a great deal from the AI revolution in health care — from more meaningful interactions to more personalized treatments, to better outcomes throughout our health care journey. Abandoning those benefits to avoid perceived risks would be careless when we have so much to gain. Ultimately, technological progress will not stop marching forward, and if we fail to embrace it, it will leave us behind.
Mike Monteiro is chief product officer for Aspire Ventures, a private equity and venture capital firm specializing in early-stage investments.