AI, radiology and the future of work

Clever machines will make workers more productive more often than they will replace them


The Economist
Jun 9th 2018


RADIOLOGISTS, say the pessimists, will be first against the wall when the machines take over. 

Analysing medical images is a natural fit for “deep learning”, an artificial-intelligence (AI) technique which first attracted attention for its ability to teach computers to recognise objects in pictures. 

A variety of companies hope that bringing AI into the clinic will make diagnosis faster and cheaper. 

The machines may even be able to see nuances that humans cannot, assessing how risky a patient’s cancer is simply by looking at a scan.


A variety of companies hope that bringing AI into the clinic will make diagnosis faster and cheaper.


Some AI researchers think that human beings can be dispensed with entirely.


“It’s quite obvious that we should stop training radiologists,” said Geoffrey Hinton, an AI luminary, in 2016. 

In November Andrew Ng, another superstar researcher, when discussing AI’s ability to diagnose pneumonia from chest X-rays, wondered whether “radiologists should be worried about their jobs”. 

Given how widely applicable machine learning seems to be, such pronouncements are bound to alarm white-collar workers, from engineers to lawyers.


In fact the application of AI to medicine suggests that the story is more complicated


Machine learning will indeed change many fields, allowing the rapid analysis of enormous piles of data to uncover insights that people might overlook. 

But it is not about to make humans redundant. 

And radiology, the very field that is used as a cautionary tale about the robopocalypse, shows why.


One is the nature of AI itself. 


The field is suffused with hype. 

Some papers show artificial radiologists outperforming the ones in white coats (see article). Others, though, still put the humans ahead. 

The machines may eventually take an unambiguous lead. But it is important to remember that AI, for the foreseeable future, will remain “narrow”, not general. 

No human is as good at mental arithmetic as a $10 pocket calculator, but that is all the calculator can do. 

Deep learning is broader. It is a pattern-recognition technique, and patterns are everywhere in nature. 

But in the end it, too, is limited-a sort of electronic idiot-savant which excels at one particular mental task but is baffled by others. 

Instead of wondering whether AI can replace a job, it is better to ponder whether it could replace humans at a specific task.


Instead of wondering whether AI can replace a job, it is better to ponder whether it could replace humans at a specific task.


The human touch


That leads to a second reason for optimism: the nature of work. 


Most jobs involve many tasks, even if that is not always obvious to outsiders. Spreadsheets have yet to send the accountants to the dole queue, because there is more to accountancy than making columns of figures add up. 

Radiologists analyse a lot of images. But they also decide which images should be taken, confer on tricky diagnoses, discuss treatment plans with their patients, translate the conclusions of the research literature into the messy business of real-life practice, and so on. 


Radiologists analyse a lot of images. But they also decide which images should be taken, confer on tricky diagnoses, discuss treatment plans with their patients, translate the conclusions of the research literature into the messy business of real-life practice, and so on.

Handing one of those tasks to a computerised helper leaves radiologists not with a redundancy cheque, but with more time to focus on other parts of their jobs-often the rewarding ones.


A third reason for optimism is that automation should also encourage demand. 


Even in the rich world, radiology is expensive. If machines can make it more efficient, then the price should come down, allowing its benefits to be spread more widely and opening up entire new applications for medical imaging.

In the Industrial Revolution the number of weavers rose as the work became more automated. Improved efficiency led to higher production, lower prices and thus more demand for the tasks that the machines could not perform. 

Medicine itself provides a more recent example. “Expert systems” were the exciting new AI technology of the 1970s and 1980s. They eventually made their way into hospitals as, for instance, automated diagnostic aids. 

That has been a boon, letting nurses-or even patients-undertake procedures that might previously have required a doctor.

Medicine itself provides a more recent example. “Expert systems” were the exciting new AI technology of the 1970s and 1980s

They eventually made their way into hospitals as, for instance, automated diagnostic aids.

That has been a boon, letting nurses-or even patients-undertake procedures that might previously have required a doctor.


No one knows how sweeping the long-term effects of AI on employment will be. But experience suggests that technological change takes longer than people think. 


Factory-owners took decades to exploit the full advantages of electricity over steam. 

Even now, the computer revolution in the office remains unfinished. 

Big tech firms such as Google, Facebook and Alibaba have the resources and the in-house expertise to begin making use of AI rapidly. 

Most other companies will proceed more slowly, especially in tightly regulated areas like medicine. 

If you happen to be training for a career in radiology-or anything else that cannot be broken down into a few easily automated steps-it is probably safe to carry on.

This article appeared in the Leaders section of the print edition under the headline “Images aren’t everything”


Originally published at https://www.economist.com on June 7, 2018.

Total
0
Shares
Deixe um comentário

O seu endereço de e-mail não será publicado. Campos obrigatórios são marcados com *

Related Posts

Subscribe

PortugueseSpanishEnglish
Total
0
Share