MARKED REDUCTION IN HALLUCINATION RATES WITH GPT-5: A POSITIVE DEVELOPMENT FOR MEDICAL AND SCIENTIFIC WRITING
- Author:
Serkan Polat
1
;
Bülent Alyanak
2
;
Burak Tayyip Dede
3
;
Mustafa Hüseyin Temel
3
;
Mustafa Turgut Yıldızgören
4
;
Fatih Bağcıer
5
Author Information
- Publication Type:Letter
- Keywords: Large language models; Artificial intelligence; Hallucinations; Medical writing; Publishing
- From:Malaysian Family Physician 2025;20(1):1-2
- CountryMalaysia
- Language:English
- Abstract: MARKED REDUCTION IN HALLUCINATION RATES WITH GPT-5: A POSITIVE DEVELOPMENT FOR MEDICAL AND SCIENTIFIC WRITING:We read with great interest the letter by Jamaluddin et al., which highlighted hallucination as a key challenge in artificial intelligence (AI)-generated writing, particularly in academic and clinical contexts.1 Since the publication of their correspondence, developments in large language models (LLMs) have progressed rapidly, with newer systems demonstrating measurable improvements in accuracy. In light of recent performance data for GPT-5, it is timely to revisit this discussion – not to reiterate the risks alone but to acknowledge the notable reduction in hallucination rates and to consider its positive implications for medical and scientific writing.
- Full text:2026110419035120326Marked reduction in hallucination rates.pdf
