#741Computer Science & AIFront Page6 January 2026
The Evolutionary Case for Shrinking AI in Ophthalmology
A recent comparative study reveals that small, specialised language models can match the accuracy of ChatGPT-4.0 when answering glaucoma questions. While both models struggled with readability, the findings suggest that privacy-focused, local AI may be the superior clinical choice.
By Faneli, Scherer, Muralidhar, Guerreiro-Filho, Beniz, Vilasboas-Campos, Costa, Jammal, Medeiros