Doctors Are Confident That AI Won't Replace Them
Doctors Are Confident That AI Won't Replace Them
What kind of world would we live in if AI took over our jobs? Doctors aren't worried.
Join the DZone community and get the full member experience.Join For Free
Bias comes in a variety of forms, all of them potentially damaging to the efficacy of your ML algorithm. Read how Alegion's Chief Data Scientist discusses the source of most headlines about AI failures here.
Companies such as DeepMind and IBM have been at the forefront of introducing AI into healthcare, with their smart algorithms proving adept at finding patterns in large quantities of data that enable the machines to make accurate predictions in the diagnosis of a range of conditions. There have been projects to provide more accurate and earlier diagnoses for mental health, dementia, Parkinson's, skin cancer, Alzheimer's, arthritis and, well, you get the picture.
The reporting of many of these projects has been accompanied by breathless claims that doctors work will soon be automated, with these claims seeming to be supported by the growing number of AI-based triage systems on the market that claim to be able to accurately diagnose patients after hearing of their various symptoms.
One might imagine that given this technological onslaught, doctors are a bit edgy about their long-term prospects, but that doesn't appear to be the case at all. Indeed, a recent study set out to ask doctors operating in primary care across the United Kingdom how they felt about AI technologies and whether they believed their job was at risk, and the answer was an overwhelming no.
Indeed, the doctors thought that the new AI technologies wouldn't be able to replace them on any of the six key medical tasks that make up their work, with the honorable exception of paperwork, which as doctors regularly cite paperwork as a reason for burnout, medical errors and ultimately leaving the profession, is perhaps not that surprising.
The researchers believe their findings highlight a rift between the medical community and the AI community, and they hope that their work will go at least some way towards smoothing over that divide. What's more, however, with AI slowly being introduced into healthcare systems around the world, they are concerned that the skepticism within the medical community about the validity of the technology could lead to inevitable conflict.
"We need a medical community that is fully engaged in critical debates about the ethics and regulation of AI in healthcare," they explain. "We need an enlightened community, but changing mindsets is always a difficult thing to do."
The survey asked doctors to rate whether they thought either current or future AI-based technology would replace them in six key tasks:
- Analyzing patient information to form diagnoses
- Analyzing patient information to reach prognoses
- Evaluating when to refer patients to colleagues
- Constructing personalized treatment plans
- Showing empathy to patients
- Creating documentation about patients
On a five-point scale that ranges from extremely likely to extremely unlikely, the doctors overwhelmingly erred towards unlikely for all of the six tasks except providing support with documentation, with perhaps understandably the delivery of empathetic care the task seen as most unlikely to be automated.
The findings support a general and growing consensus that early iterations of AI will largely support the work of people rather than replace them, and that it will perform best when providing efficient and effective predictions. So in medicine, AI will augment what they do rather than replace them.
A Degree of Confidence
It's perhaps worth saying that such confidence in the face of technology is not confined to the medical profession. A recent study from Massey University's School of Management asked people in New Zealand whether they were concerned about technology taking their job.
Despite the various doomsday predictions about the impact of technology on jobs, the study found that 87 percent of workers thought their job would largely be safe from automation.
"It was interesting that those who most strongly denied the possibility of a machine doing their job were often from the sectors most at risk, like checkout operators, drivers, and analysts. These are all areas where we can already see technology having an impact," the authors explain.
Which suggests that their confidence is perhaps misplaced and down more to ignorance of the technology than any realistic perception of the threat posed by technology. Does the same apply to medics? It's hard to say, but a recent study undertaken by Wharton in relation to the launch of the Data Literacy Index found that the healthcare sector scored the lowest in terms of its data literacy.
This suggestion was supported by a recent study from PwC, which examined the 'digital IQ' of various industries. The study revealed that just 21% of executives in the healthcare sector thought that they were at any risk of digital disruption, and this was despite just 3% of healthcare providers spending the bulk of their digital investments on 'disruptive' technologies such as AI.
"In PwC's 2018 Digital IQ research findings, we found that almost half of the health services companies surveyed spend 6-10 percent on digital investments, with overall top industry performers spending over 10 percent. When looking at how funds are allocated throughout the industry, we found that only 3 percent of health services companies are currently investing in emerging technologies and artificial intelligence, emphasizing that the industry as a whole doesn't prioritize digital transformation as much as other industries. There is far more concentration on investing in the customer experience realm and the human resources space than in new and emerging technology," David Clarke, Global Chief Experience Officer at PwC, told me recently.
Slow Pace of Change
Of course, as well as perhaps a low level of digital literacy, the sector is also renowned for the slow pace of technological change. For instance, even a technology as proven as video consultations have taken the best part of a decade to achieve minimal results.
A study exploring the rollout of video consultations in Norway highlights the glacial pace of change. Despite the country having a five-year plan to provide digital consultations way back in 2009, and providing generous funding and strong policy support to achieve this, by 2013, just 2 percent of outpatient consultations were conducted by video link. This was despite 75 percent of hospitals signing up to the plan.
Similar frustrations have emerged throughout the world, as healthcare providers struggle to adapt to the changing world around them. In such a slow-paced environment, it's perhaps not that surprising, therefore, that doctors feel pretty confident that the winds of change will not be blowing their way.
Published at DZone with permission of Adi Gaskell , DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.