A recent report from Accenture concludes that healthcare executives are enthusiastic about “intelligence technologies” such as AI and IoT. It also suggests, however, that health organizations will need to add new capabilities to be sure they can manage these technologies responsibly.
The report, based on a survey of 100 health executives, found that 77% of respondents expect to invest in IoT and smart sensors and that 53% expect to invest in AI systems. Presumably, they expect these technologies to offer benefits more quickly.
Why the gap in adoption? The truth is that healthcare leaders haven’t yet gotten their arms around AI just yet. While IoT and smart sensor technology can boost the flexibility and “judgment” of enterprise systems, AI arguably has the potential to be far more flexible and wide-reaching — and ultimately less than predictable.
This unpredictability makes AI investment a bit trickier to implement than other emerging technologies. Just over four-fifths of health leaders said they were not prepared to explain AI-based conclusions to their internal stakeholders nor outsiders.
To address this deficit, 73% said they plan to develop internal ethical standards for AI to make sure these systems can act responsibly. Before that, they’ll need to determine what “acting responsibly” actually means — and as far as I know there are no accepted guidelines for developing such standards. (They might want to start off by reviewing Google’s ethical principles for AI use here.)
Adding AI to the enterprise IT mix could also wreak havoc. I for one was surprised to read that almost one-fourth of respondents said that they had been the target of adversarial AI behaviors, including falsified location data or bot fraud. (This stat blew my mind. Why haven’t we heard more about these “adversarial behaviors” and what are they?)
This certainly adds another element of uncertainty for CIOs interested in AI investments. While AI technologies can’t “think” in the traditional sense, they can create a range of problems previous-gen technology couldn’t.
This is part of a larger picture in which health organizations aren’t sure if their data has been corrupted. In fact, 86% of health execs said they hadn’t yet invested in technologies which could verify their data sources. Adding AI to the mix could potentially compound these problems, as it might create a cascade in which the AI then draws false inferences and takes inappropriate actions.
Meanwhile, respondents were excited about blockchain and smart contracts technology, with 91% reporting that they believed it would be a critical tool for supporting frictionless businesses over the next three years. All told, expect to see IoT and blockchain investments right away, with AI lagging until health IT leaders can teach it to play nicely.