What is uniqueness neglect?
There is a tendency not to trust the forecast of statistical models even if they are superior to human decision-making (Meehl 1954, in Longoni et al. 2019). In other words, even though statistical models outperform human intui-tion when making predictions, individuals still prefer human expertise when mak-ing their own decisions. Currently, there is a reluctance to trust the decision of an artificial intelligence system in the medical field even if the system is not statistical in nature and has proven superior to medical experts (Longoni et al. 2019).
Yet, artificial intelligence is revolutionising the field of healthcare. AI systems have proven effective in the diagnosis of heart disease, in providing medical advice as part of National Health Services as well as the detection of skin cancer (Longoni et al. 2019). Even though all these systems perform with expert level accuracy, there is still a reluctance to accept decision making of an AI system. Longoni et al. (2019, p.631) write
"We propose that consumers may be more reluctant to utilize healthcare delivered by AI providers than healthcare delivered by comparable human providers, even when they are explicitly informed about the performance of the providers. We further propose that this occurs because the prospect of being cared for by an automated provider evokes a concern that one’s unique characteristics, circumstances, and symptoms will be neglect-ed. We refer to this concern as uniqueness neglect. We argue that uniqueness neglect emerges out of a mismatch between two fundamental beliefs. Whereas consumers view themselves as unique and different from others …, consumers view machines as capable of operating only in a standardised and rote manner that treats every case the same way."
It is easy for a clinical psychologist to relate to this phenomenon of uniqueness neglect. Frequently, clients present with concerns in daily clinical practice that they describe as completely out of the ordinary. Clients may even apologise for presenting with concerns because of their perceived uniqueness and the view that a certain presentation separates the individual from the rest of society. Once the clinician has conducted the assessment, it may turn out that the core of the concern is related to depression and anxiety, i.e. psychological problems that are very frequent in western societies.
If it is part of the cognitive self model that individuals view their own behav-iour and thought processes as unique, how can a statistical model or AI system then accommodate for this uniqueness concern?
The answer is of course that people who live in the same society are exposed to the same stressors and risk factors that contribute to mental health problems. People share cultural norms and expectations as well as similar family and work environments. As a result, individuals tend to develop similar mental health prob-lems. The cognitive dysfunctions and unhelpful thinking styles that can cause depression are a very good example: many people catastrophise or personalise their problems or use emotional reasoning. As a result, there is an overlap in psychiatric symptoms and true uniqueness is rare.
However, if the uniqueness neglect is an obstacle to the adoption of AI systems in the medical sector, then the question is
"how medical AI … integrated into healthcare delivery may profoundly influence patients’ satisfaction with medical services, a metric that determines key outcomes for healthcare institutions" (Centers for Medicare and Medicaid Services 2019).
Diederich J, The Psychology of Artificial Superintelligence. Springer Nature Switzerland AG 2021, ISBN 978-3-030-71841-1, DOI https://doi.org/10.1007/978-3-030-71842-8
Longoni C, Bonezzi A, Morewedge CK, Resistance to Medical Artificial Intelli-gence. Journal of Consumer Research, Vol. 46 (2019) 629-650.
uniqueness
What is uniqueness neglect?
There is a tendency not to trust the forecast of statistical models even if they are superior to human decision-making (Meehl 1954, in Longoni et al. 2019). In other words, even though statistical models outperform human intui-tion when making predictions, individuals still prefer human expertise when mak-ing their own decisions. Currently, there is a reluctance to trust the decision of an artificial intelligence system in the medical field even if the system is not statistical in nature and has proven superior to medical experts (Longoni et al. 2019).
Yet, artificial intelligence is revolutionising the field of healthcare. AI systems have proven effective in the diagnosis of heart disease, in providing medical advice as part of National Health Services as well as the detection of skin cancer (Longoni et al. 2019). Even though all these systems perform with expert level accuracy, there is still a reluctance to accept decision making of an AI system. Longoni et al. (2019, p.631) write
"We propose that consumers may be more reluctant to utilize healthcare delivered by AI providers than healthcare delivered by comparable human providers, even when they are explicitly informed about the performance of the providers. We further propose that this occurs because the prospect of being cared for by an automated provider evokes a concern that one’s unique characteristics, circumstances, and symptoms will be neglect-ed. We refer to this concern as uniqueness neglect. We argue that uniqueness neglect emerges out of a mismatch between two fundamental beliefs. Whereas consumers view themselves as unique and different from others …, consumers view machines as capable of operating only in a standardised and rote manner that treats every case the same way."
It is easy for a clinical psychologist to relate to this phenomenon of uniqueness neglect. Frequently, clients present with concerns in daily clinical practice that they describe as completely out of the ordinary. Clients may even apologise for presenting with concerns because of their perceived uniqueness and the view that a certain presentation separates the individual from the rest of society. Once the clinician has conducted the assessment, it may turn out that the core of the concern is related to depression and anxiety, i.e. psychological problems that are very frequent in western societies. If it is part of the cognitive self model that individuals view their own behav-iour and thought processes as unique, how can a statistical model or AI system then accommodate for this uniqueness concern?
The answer is of course that people who live in the same society are exposed to the same stressors and risk factors that contribute to mental health problems. People share cultural norms and expectations as well as similar family and work environments. As a result, individuals tend to develop similar mental health prob-lems. The cognitive dysfunctions and unhelpful thinking styles that can cause depression are a very good example: many people catastrophise or personalise their problems or use emotional reasoning. As a result, there is an overlap in psychiatric symptoms and true uniqueness is rare. However, if the uniqueness neglect is an obstacle to the adoption of AI systems in the medical sector, then the question is
"how medical AI … integrated into healthcare delivery may profoundly influence patients’ satisfaction with medical services, a metric that determines key outcomes for healthcare institutions" (Centers for Medicare and Medicaid Services 2019).
Diederich J, The Psychology of Artificial Superintelligence. Springer Nature Switzerland AG 2021, ISBN 978-3-030-71841-1, DOI https://doi.org/10.1007/978-3-030-71842-8
Longoni C, Bonezzi A, Morewedge CK, Resistance to Medical Artificial Intelli-gence. Journal of Consumer Research, Vol. 46 (2019) 629-650.