Az LLM-ek pontosabbak a doktoroknál és a neurológusoknál.
Large language models can predict the results of proposed neuroscience studies more accurately than human experts, finds a study led by UCL (University College London) researchers.
...
"All of the LLMs outperformed the neuroscientists, with the LLMs averaging 81% accuracy and the humans averaging 63% accuracy. Even when the study team restricted the human responses to only those with the highest degree of expertise for a given domain of neuroscience (based on self-reported expertise), the accuracy of the neuroscientists still fell short of the LLMs, at 66%."
Ehhez kapcsolódóan:
Study Finds ChatGPT Outperforms Doctors at Diagnosing Illness
In a surprising small-scale study, ChatGPT-4 outperformed doctors at diagnosing medical cases, even when those doctors had access to the same chatbot. Published in JAMA Network Open, the study tested 50 physicians on six challenging medical cases. The chatbot scored 90%, compared to 76% for doctors using ChatGPT and 74% for those relying only on conventional resources.
The findings highlight three critical issues:
1. Human Bias: Many doctors clung to their initial diagnoses, even when the chatbot suggested alternatives with better reasoning.
2. Underutilization: Most physicians used ChatGPT for targeted questions, failing to exploit its ability to analyze entire case histories comprehensively.
3. Trust Gap: Despite ChatGPT’s superior performance, skepticism remains about integrating A.I. into clinical workflows.Historically, A.I. has struggled to find its place in diagnostics, hindered by usability and trust issues. Unlike earlier systems, ChatGPT doesn’t mimic human diagnostic reasoning. Instead, it leverages language prediction to deliver fast, accurate insights. However, its success depends on how well users understand and utilize its capabilities.
• ChatGPT scored 90%, while doctors scored 74%-76% in diagnosing cases.
• Physicians often resisted chatbot insights that contradicted their initial beliefs.
• Only a few doctors maximized ChatGPT’s potential by submitting full case histories.
• Study underscores the need for better A.I. training and adoption among medical professionals.