Pandemic-Driven AI Adoption is Remaking Industries, Creating an Uncertain Future
Here's a look at how pandemic-accelerated AI adoption is remaking some businesses and what effects that might have in the long run.
Join the DZone community and get the full member experience.Join For Free
Throughout 2020, the COVID-19 pandemic has dominated headlines around the world, and rightly so. And for the most part, discussions about how it intersected with the technology world have centered on things like contact tracing and epidemiology research topics. But as the pandemic has continued to batter populations and economies everywhere, there's growing evidence that it's becoming a major digital change driver across a wide variety of industries, too.
One after another, surveys continue to uncover an acceleration of the ongoing trend toward AI adoption in multiple fields. One, in particular, found that 68% of businesses have increased their spending on AI technology as a part of their pandemic response. When all is said and done, we could be witnessing a fundamental shift that will leave the economic landscape looking very different than it did less than a year ago.
But to understand what that means and what specific effects it could produce, it's necessary to dig into some of the available data to see exactly what technology's being adopted and where it's being put to use. Here's a look at how pandemic-accelerated AI adoption is remaking some businesses in real-time and what effects that might have in the long run.
A Healthcare Evolution
Any way you look at it, the healthcare industry is at the top of the list when it comes to AI adoption, and technology adoption more broadly. Some of the movement toward AI in healthcare predates the pandemic, to be sure, but it has also radically altered timelines for AI adoption in a variety of use cases.
For example, Mumbai-based Qure.ai was engaged in clinical trials at the Royal Bolton Hospital, an NHS-run hospital in the UK before the pandemic began. Their AI-powered chest X-ray evaluation system – called qXR – was to undergo efficacy evaluations over six months in 2020. Then, the COVID-19 pandemic changed things.
Suddenly, the hospital, looking for ways to ramp up COVID-19 diagnosis capacity, turned to the developers for help. In mere weeks, the qXR system was pressed into service as a frontline prescreening tool to detect the telltale lung damage produced by the rampaging virus. Shortly thereafter, similar applications began regular use throughout hospitals in France.
And that situation wasn't an outlier. Other healthcare organizations also rolled out conversational AI solutions to create at-home triage capabilities to stem the tide of low-risk patients flooding into hospitals. The technology proved vital in keeping several major hospital chains and government healthcare systems afloat when they were close to their breaking point earlier this year.
Healthcare providers weren't alone in their mad scramble to leverage AI to keep working through the pandemic. Businesses of all kinds turned to machine learning and automation technology to reimagine processes while employees were confined in their homes. In one very visible case, pharmaceutical company Takeda rolled out RPA to help speed up clinical trials for potential COVID-19 treatments.
But the tools worked so well that they then created a plan to train thousands of their employees to create their own bots to automate routine work tasks. And in the years ahead they expect to use their early RPA experience to transition into full-blown AI task automation in the coming years.
And then there were businesses like Juniper Networks, which put its Mist AI platform to work to build a proximity-based worker tracking system. It uses Bluetooth Low Energy (BLE) tags on employees' ID badges to track employee movements within company spaces in real-time. The system was key to the company getting its most critical engineering teams back to work with minimal risk of getting infected with COVID-19.
Their efforts, and efforts like them, are transforming the way offices function. They're also introducing new elements and capabilities that could create new efficiencies, render certain jobs obsolete, and even introduce unforeseen risks to the businesses themselves. And because of the accelerated adoption curve made both possible – and necessary – by the pandemic, there's a good chance that businesses haven't taken a pause to assess the situation.
What Pandemic-Driven AI Adoption Will Mean
When evaluating what the ultimate outcome of this pandemic-driven surge in AI adoption will be, it's tempting to rely on some big pre-pandemic assumptions to find a starting point. More specifically, it's safe to assume that some of the outcomes that have already been predicted are at least likely to take place, albeit sooner than expected.
For example, the sudden move toward RPA and even AI-powered medical task automation is all but certain to accelerate job losses in more than a few industries. But it's also likely to create some headwinds for the global outsourcing market since it's dependent on some of the jobs that are most likely to become redundant.
But more importantly, the rush into AI is likely to lead to unforeseen consequences simply because so many development and testing timelines were rushed to accommodate unexpected business needs. For example, there were good reasons that the Royal Bolton Hospital was planning to run such a lengthy trial on Qure.ai's technology.
It's because, in a medical setting, inaccurate AI-driven diagnoses can cost lives. And that's not all. There's also great concern that AI and ML systems can include biases that are both hard to detect and even harder to correct. And that raises the possibility of disparate health outcomes for certain groups compared to others.
And beyond that, rushed technology deployments, in general, don't tend to work out well. When dealing with something as complex as AI, that goes double. So that means some of the companies who jumped head-first into the AI pool could be sitting on something of a high-tech time bomb that they can neither see nor address.
What Comes Next
The early returns don't seem promising, either. A recent survey of companies that have integrated AI solutions into their workflows found that a scant 11% reported deriving 'sizable' benefits from doing so. And when you consider many of the respondents to the survey were businesses who bought into AI before the pandemic hit (which presumably means they took their time with it), it seems exceedingly unlikely that pandemic-era rushed AI adoptions will turn out to be long-term successes without some major retooling along the way.
None of this, of course, means that the pandemic won't prove to be a transformative force that moved AI into the economic mainstream much faster than it otherwise would have gotten there. In fact, simple inertia should be enough to make that a fait accompli. After all, large organizations don't tend to throw away significant investments without making every effort to get their money's worth out of them.
And it's also possible that the structural changes in the workforce that the sudden rush to automation will create will make it so that companies have little choice but to make their AI solutions work for them. After a certain amount of time, displaced workers will retrain and move on, making it impossible to go back to the way things were.
No matter what happens next, though, what seems certain is that there won't be many businesses or industries left that haven't in some way moved closer to adopting AI technology by the time the pandemic ends. The only real question is how well they'll manage the technology and how quickly AI developers can learn from these real-world uses and make their technologies better and more reliable. But that's a topic for another discussion altogether.
Opinions expressed by DZone contributors are their own.