MD+DI Online is part of the Informa Markets Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Will AI Be a Savior or a Demon for Medtech?

The Terminator franchise popularized the notion of artificial intelligence becoming a threat to humanity. Terminator image from Flickr user Dick Thomas Johnson.

So how long before artificial intelligence (AI) transcends the realm of gadgetry and makes its way into the workforce--including the medtech workforce? Already, AI is being used to design websites. Could robots design medical devices, say, a decade from now?

It might not be as crazy as it sounds. In any event, the medical field is becoming increasingly enamored with with the potential of AI to reboot the field. The big question is whether AI may inadvertently become a big problem for humanity someday, rather than a savior.

Before considering that question, let's take a look at how far AI has come even recently. IBM, for example, is now touting its Watson supercomputer as an invaluable AI version of a physician's assistant. The computer, which famously trounced human opponents in Jeopardy! in 2011, was 24 times faster and 90% smaller in 2014 than it was three years earlier. The system could potentially be especially valuable in medicine, as it can read huge troves of medical journals within minutes while never getting bored. Last year, Watson could scan 10 billion files in 45 minutes. It can also understand natural human language. These capabilities could enable it to help diagnose a patient based on unthinkable amounts of data--giving a list of possible diagnoses with confidence levels for each scenario.

At the same time, the word "assistant" is important. IBM says that doctors will still have to make the final call, though it may be hard to disagree with a supercomputer's advice--especially considering the supercomputer is always getting smarter and can learn from its mistakes.

A considerable number of physicians were dismayed when the entrepreneur and investor Vinod Khosla first predicted that algorithms could one day replace 80% of what doctors do back in 2012. He's still convinced that AI will ultimately lead to big changes in medicine. "Inevitably, over 20 years, the majority of physicians' diagnostic, prescription and monitoring functions will be replaced by smart hardware, software and testing," he explained in a TechCrunch article last year.

Advanced artificial intelligence has already replaced countless jobs from data entry and processing, to manufacturing devices and various goods. And artificially intelligent system are starting to be able to perform tasks that were once only skilled people could do: a growing number of articles, for instance, are now written by robots. And artificial intelligence is being used to help facilitate legal research.  

It might not be long before self-driving cars and trucks come along. And once self-driving trucks hit the highways, they could replace drivers pulling in Teamsters wages. Already, Uber has emerged as a threat for conventional cab drivers. What would happen if Uber was powered by a network of self-driving cars? Or, for medical device companies, what if robots could be used to automate the submission of at least some regulatory paperwork? Or if it could assist engineers with design for manufacturability analysis?

Recently Murray Shanahan, a professor of cognitive robotics at Imperial College, spoke in an interview with Wired about the reach of AI technology extending into the workplace around us. According to Shanahan, artificial intelligence will begin to have a significant effect on society over the next five to 10 years.

"It's hard to predict exactly what's going on," Shanahan said. "But we can be pretty sure that these technologies are going to impact society quite a bit."

This idea has been reinforced by two researchers from MIT who argue that we're entering a "second machine age," a time where the rate of change brought on by digital technologies could accelerate at such alarming rates, that it could leave millions of medium and low-skilled workers out of a job, and that could just be the beginning. With recent advances in perception technology, the range of machine functionality has become even more diverse. Computers are beginning to understand what we say, before analyzing data and executing tasks, much like a secretary or assistant would do.

Last year, at the MIT AeroAstro Centennial Symposium, Elon Musk likened using artificial intelligence to summoning demons. "In all those stories where there's the guy with the pentagram and the holy water, it's like - yeah, he's sure he can control the demon. Doesn't work out," explains Musk, who also said that artificial intelligence is our biggest existential threat.

Even Bill Gates recently discussed the dangers of surging artificial intelligence, when participating in a Q&A session on the popular social website known as Reddit. When asked about the potential threat of super artificial intelligence on our society as a whole, his response was rather cautionary.

"I am in the camp that is concerned about super intelligence," Gates wrote. "First the machines will do a lot of jobs for us and not be super intelligent. That should be positive if we manage it well. A few decades after that, though, the intelligence is strong enough to be a concern. I agree with Elon Musk and some others on this and don't understand why some people are not concerned."

While that kind of interpretation may sound somewhat alarming, many have moved to cool the idea of artificial intelligence ever becoming a serious threat to the existence of humanity. Eric Horvitz, director of Microsoft Research's Redmond Lab, remarked in an interview with the BBC that much of the doomsday declarations surrounding AI have been overblown.

"There have been concerns about the long-term prospect that we lose control of certain kinds of intelligences," Horvitz said to the BBC. "I fundamentally don't think that's going to happen. I think that we will be very proactive in terms of how we field AI systems, and that in the end, we'll be able to get incredible benefits from machine intelligence in all realms of life. From science, to education, to economics, to daily life."

While the idea of machines overtaking humans remains more of a science fiction narrative, the idea of AI increasing its role in our everyday life, and extending into the workforce, becomes more and more realistic by the day. Elon Musk is investing in artificial intelligence--not so much to profit from it, but to keep tabs on it as the technology progresses.

Horvitz also noted in his interview that over a quarter of Microsoft's attention and resources at Microsoft Research are focused on artificial intelligence, proving that AI isn't just a nerdy gimmick, but a very real growing presence that's worth keeping an eye on.

Refresh your medical device industry knowledge at MD&M West, in Anaheim, CA, February 10-12, 2015.

Kristopher Sturgis is a contributor to Qmed and MPMN.

Like what you're reading? Subscribe to our daily e-newsletter.

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.