Aging AI Chatbots Show Signs of Cognitive Decline in Dementia Test – Casson Living – World News, Breaking News, International News

Aging AI Chatbots Show Signs of Cognitive Decline in Dementia Test – Casson Living – World News, Breaking News, International News


Aging AI Chatbots Show Signs of Cognitive Decline in Dementia Test

In recent times, AI models have been under scrutiny for their peculiar behaviors, with new research suggesting they may even display signs of cognitive decline. A study published in The BMJ has found that several leading chatbots in the tech industry are showing symptoms akin to mild cognitive impairment, particularly as they age.

The objective of the study was to challenge the prevailing belief that AI technology has matured sufficiently for applications in the medical field, especially for diagnostic tasks. Researchers discovered that prominent chatbots like GPT-4 and Gemini versions 1.0 and 1.5 struggled with essential tasks involving visuospatial skills and executive functions, such as accurately drawing a clock or remembering sequences of words. These limitations raise serious questions about the dependability of AI in medical diagnostics.

Interestingly, when evaluated using the Montreal Cognitive Assessment, GPT-4 achieved the highest score among the tested chatbots, while the Gemini models lagged behind. Despite performing well in language-related tasks, all chatbots exhibited a notable deficiency in empathy—a significant indicator of frontotemporal dementia.

The researchers urge caution against anthropomorphizing AI systems, warning that they should not be held to human standards. Yet, they contend that if companies are marketing these AI models as sentient entities, they must be accountable for their cognitive performance.

Ultimately, the findings of this study indicate that AI models are not poised to take over the roles of human doctors in the near future. Furthermore, healthcare practitioners may soon find themselves dealing with virtual patients—AI systems that exhibit cognitive impairments. This research underscores the limitations of current AI technology and prompts a reevaluation of the industry’s claims regarding the capabilities of these sophisticated chatbots.