As new technologies enter a growing number of fields, new research finds that our general faith in technology has a big impact on our willingness to use new tools, with a knock-on effect on the speed of adoption.
Trust building exercises are a staple of team workshops, and whilst the merits of hoping Francis from accounts will catch you as you fall backwards are somewhat dubious, the merits of trust in the successful functioning of any team is certainly not. Whether it’s supporting innovation, knowledge sharing or general group cohesion, trust is key.
It’s perhaps no surprise therefore that trust also has a big role to play in how we interact with technology, especially when that technology has autonomous capabilities. For instance, recent research from the University of Hawaii highlighted how our relative trust in the power of technology has an understandably large role to play in our willingness to use the autonomous driving functions in the latest Tesla vehicles.
A triumvirate of studies from Penn State’s S. Shyam Sundar reveals that this is consistent across multiple sectors, each of which depends upon faith between consumer and supplier is paramount. As with the team from Hawaii, he began with an exploration of autonomous driving technology, and especially what factors underpin our willingness to accept the technology in our lives.
One of the more commonly cited examples of superiority bias is that most of us think we’re above average drivers, and while this has obvious implications for things like road safety, the authors also found that it can hamper the adoption of driverless technology. The study showed that people who strongly believed that self-driving cars were more capable than they themselves were much more likely to be in favour of the technology, both for themselves and indeed for other road users.
While this is perhaps not surprising, what is interesting is that this need for confidence in the capability of the technology was far more important in terms of acceptance than all other factors, whether it was the coolness of the technology or simply loving technology more generally. As well as impacting adoption, Sundar believes this varying degree of acceptance could also affect how designers construct the vehicles themselves, with the potential for the steering wheel to be removed from the interior and replaced with features that enhance the agency and convenience for the passenger.
Trusting the robodoc
Just as we have a pretty high opinion of our driving capabilities, we also have tremendous faith in medical professionals. Indeed, the latest edition of the annual Ipsos Veracity Index revealed that 96% of people trust nurses to tell the truth, and 92% trust doctors to do so. Any technology attempting to muscle in on this turf has to be pretty impressive, therefore. The acceptance of medtech devices was the second domain analysed by Sundar, and as before, he assessed whether our faith in technology led to a greater acceptance of new medtech applications providing us with advice and diagnoses about our health and wellbeing.
As before, the study found that people who were more comfortable with the latest technologies were more comfortable with accepting the latest medtech devices into their lives, both because they were more confident in the capabilities of the technology, and in their personal ability to use the technology.
This finding is interesting in the context of the end user, but also in the context of clinicians themselves, who are likely to be consumers of this technology as it augments their clinical practice. Previous research has shown that doctors are extremely confident that AI will never replace 5 of the 6 core tasks that make up their work (note taking being the honourable exception), and there are long-standing accusations that doctors are not the tech-savvy people Sundar believes will be the early adopters of AI-driven technology in medicine.
Central to Sundar’s work was something known as ‘machine heuristics’, which in layman’s terms refers to the default thinking we tend to deploy when we think about machines and technology. Participants in his research were asked to state their level of agreement with statements such as “When machines, rather than humans, complete a task, the results are more accurate.”
While doctors as a profession were not singled out for analysis in the report, the research mentioned earlier does suggest they would respond in the negative to such a statement, which Sundar reveals makes them less positive in their attitude towards AI-driven technologies in healthcare than those with a higher belief in the machine heuristic.
This boost was especially strong when the technical competence of the volunteer was high, but again, this is an area the healthcare sector doesn’t score highly on. The Data Literacy Index seeks to chart data literacy across the economy, and they found that data literacy in healthcare was the lowest of the 14 sectors covered, suggesting that medics have neither strong technical skills or a strong belief that technology can benefit their work.
Safe and secure
The final domain explored by Sundar was around the ability of technology to be trusted with sensitive and personal information. It’s a topical theme as there has been a groundswell of concern about digital assistants, such as Amazon’s Alexa, and just what they listen to every day, and what happens to that data.
Interestingly, Sundar’s study suggests that consumers are generally pretty relaxed by the prospect of technology guarding our sensitive information. Indeed, it found that people are usually more trusting of technology than they would be of humans, who many of us believe are capable of the kind of dishonesty and fraud that machines are not.
What’s more, volunteers revealed that they don’t believe technology is capable of gossiping, lying or any number of negative behaviours that are very much a part of the human DNA. That’s not to say, of course, that technology will always act appropriately, and there exists the potential for unintended consequences to even the most ethically designed machines but for now, at least, the standard machine heuristic for many people is that technology acts in a more honest and trustworthy way than humans.
It’s certainly an interesting triumvirate of studies, and meshes nicely with the call I highlighted in my previous article from the MIT Media Lab for a new body of research to explore and understand the interactions between humans and technology so that we can ensure the two work effectively, efficiently and ethically together. It’s an early stage in this exploration, but these findings provide an interesting indication of the direction of travel.
New perspectives required to create AI fit for humanity
Home is where the AI is
To enable comments sign up for a Disqus account and enter your Disqus shortname in the Articulate node settings.