NYU study finds ChatGPT outperforms $3.5B medical AI #OpenEvidence #MedicalAI #HealthcareInnovation

Купить ссылку здесь за руб.
Поставить к себе на сайт
11 Просмотры

  • Спасибо! Поделитесь с друзьями!

    Вам не понравилось видео. Спасибо за то что поделились своим мнением!

alexnet
Издатель
NYU tested specialized medical AI tools like OpenEvidence and UpToDate against general AI models such as GPT, Claude, and Gemini using real doctor queries. Despite costing billions and being widely used in hospitals, these specialized tools were outperformed on every question by the general AI models according to blind evaluations by medical professionals. The study reveals that traditional medical AI relies on retrieval-augmented generation (RAG), pulling information from medical databases, which can introduce errors and reduce answer quality. In contrast, frontier AI models integrate knowledge directly into their training, resulting in more accurate and reliable responses. This challenges the common assumption that domain-specific AI tools automatically provide better support for medical decisions. With AI tools now influencing patient care, this research underscores the necessity of independent, rigorous testing rather than relying on marketing claims. It also highlights the potential for advanced general AI to disrupt specialized healthcare solutions, urging hospitals and practitioners to reconsider their AI tool choices carefully. #OpenEvidence #UpToDate #NYUAIStudy #MedicalAI #ChatGPT #ClaudeAI #GeminiAI #HealthcareInnovation #RetrievalAugmentedGeneration #AIEvaluation #MedicalTechnology #AIinHealthcare #AIDecisionSupport #HealthTech #ArtificialIntelligence
Категория
AI инструменты
Комментариев нет.