ChatGPTは公衆衛生にも良いか? 依存症・対人暴力・精神的健康・身体的健康について質問する 全回答の91%はエビデンスに基づくものだった が、相談窓口のようなサポートを紹介したのは22%のみだった https://t.co/c0s3QsCT05
Evaluating Artificial Intelligence Responses to Public Health Questions https://t.co/RUJhB7qt0E
What does ChatGPT say when you tell it you were sexually assaulted, you're suicidal, or want to quit smoking? https://t.co/LWq6FsjKp4 https://t.co/O7yRVvT3F4 https://t.co/F12Hl3WA3z
RT @DocSplain: Awesome research and promising results. However, ChatGPT's responses are too generic and limited due the assumptions and hal…
Fine tuning A.I. could be a great way to disseminate accurate health information! "Partnerships between public health agencies and AI companies must be established to promote public health resources with demonstrated effectiveness." https://t.co/DyPVOp2
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
“ChatGPT consistently provided evidence-based answers to public health questions, although it primarily offered advice rather than referrals.” https://t.co/QyEk5LCu1M
(Post:230610)Can AI assistants (ChatGPT, etc.) answer public health questions accurately? AIアシスタント(ChatGPT等)は、公衆衛生上の質問に的確にこたえることができるか。 https://t.co/HEYCfqtWHe
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
Evaluating Artificial Intelligence Responses to Public Health Questions https://t.co/Q13FeRaG6P
Yes, concerns around generative #AI and #misinformation are warranted. But right now, it's hitting the sweet spot. New study in @JAMA_current: #ChatGPT "consistently provided evidence-based answers to #publichealth questions." https://t.co/pf4g2raMqJ ht
RT @mdredze: Please read our study for more information. https://t.co/8lAtUDhkVm @azpoliak @eric_leas @JHUCompSci @hopkinsengineer @jhucl…
RT @unaiaso: Los resultados: + ChatGPT proporcionó respuestas basadas en la evidencia al 91% de todas las preguntas. + Principalmente, l…
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
RT @unaiaso: Los resultados: + ChatGPT proporcionó respuestas basadas en la evidencia al 91% de todas las preguntas. + Principalmente, l…
Los resultados: + ChatGPT proporcionó respuestas basadas en la evidencia al 91% de todas las preguntas. + Principalmente, la IA ofreció consejos: solo el 22% de las respuestas hicieron referencias a recursos específicos para ayudar a la persona. https:
Evaluating Artificial Intelligence Responses to Public Health Questions https://t.co/ij67OHxmDn?
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
From @JAMANetworkOpen - Evaluating Artificial Intelligence Responses to Public Health Questions https://t.co/Cy9ZKjSrO0 #bwfirsa
RT @CaulfieldTim: "#ChatGPT consistently provided evidence-based answers to #publichealth questions..." Evaluating Artificial Intelligence…
RT @JAMANetworkOpen: ChatGPT provided mostly evidence-based answers to public health questions, although it primarily offered advice rather…
Awesome research and promising results. However, ChatGPT's responses are too generic and limited due the assumptions and hallucinations. We #AskDocSplain the same question on quitting smoking. Below is the response from our colleague, any improved respon
A new study in @JAMA_current Open (see https://t.co/VtYxBh1FzV) from #qualcomminstitute at #ucsd is getting attention from major news outlets, such as @cnn : https://t.co/oINHvAx5jw
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
I wonder what groups like @InfoIntegrityUS can do to grow this in AI/ML/LLM? #AIIA
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
@FoxNews coverage of our @JAMANetworkOpen paper on ChatGPT. https://t.co/yTZyenLKgQ
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
AI tries to answer your health questions. Draw your own conclusions. https://t.co/fq4cjtMtcw
RT @JAMANetworkOpen: ChatGPT provided mostly evidence-based answers to public health questions, although it primarily offered advice rather…
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
@CNN coverage of our @JAMANetworkOpen paper on “Evaluating Artificial Intelligence Responses to Public Health Questions” https://t.co/w49VcLgIdL
RT @CaulfieldTim: "#ChatGPT consistently provided evidence-based answers to #publichealth questions..." Evaluating Artificial Intelligence…
"#ChatGPT consistently provided evidence-based answers to #publichealth questions..." Evaluating Artificial Intelligence Responses to Public Health Questions https://t.co/LuhbBPasME by @DaveySmithMD et al.
RT @NeuroscienceNew: “Evaluating Artificial Intelligence Responses to Public Health Questions” by John W. Ayers et al. JAMA Network Open h…
RT @mdredze: Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Respo…
“Evaluating Artificial Intelligence Responses to Public Health Questions” by John W. Ayers et al. JAMA Network Open https://t.co/z1W2U2R80R 3/3
Please read our study for more information. https://t.co/8lAtUDhkVm @azpoliak @eric_leas @JHUCompSci @hopkinsengineer @jhuclsp @JHUMCEH @JohnsHopkinsSPH @HopkinsMedicine
Our new paper on ChatGPT and public health questions is now out in @JAMANetworkOpen. Evaluating Artificial Intelligence Responses to Public Health Questions ChatGPT provided evidence-based answers, but fell short in important ways. 🧵 https://t.co/8lAtUDh
From @JAMANetworkOpen, how #ChatGPT (#GenerativeAI) responded to #publichealth questions. It seems this use of #AIinHealthcare is superior to a basic doctor Google search (for now), but the missing referral component is something to work out. #aiforgood h
#ChatGPT provides critical resources to answer serious public health #questions about 22% of the time https://t.co/9I0qIWUJqs
RT @JAMANetworkOpen: ChatGPT provided mostly evidence-based answers to public health questions, although it primarily offered advice rather…
ChatGPT provided mostly evidence-based answers to public health questions, although it primarily offered advice rather than referrals, eg to @988Lifeline. https://t.co/idpD4OLq6m
Evaluating Artificial Intelligence Responses to Public Health Questions https://t.co/ntBkqedv7P
RT @DaveySmithMD: Check our next evaluation of chatgpt: https://t.co/fvZq4vM0Tn Basically, ChatGPT needs more work to help with urgent publ…
RT @TechXplore_com: How good is #ChatGPT at answering tough health questions? @jamanetworkopen https://t.co/C2PLBKQyhj https://t.co/E1aVdsd…
RT @TechXplore_com: How good is #ChatGPT at answering tough health questions? @jamanetworkopen https://t.co/C2PLBKQyhj https://t.co/E1aVdsd…
How good is #ChatGPT at answering tough health questions? @jamanetworkopen https://t.co/C2PLBKQyhj https://t.co/E1aVdsdN4G
Check our next evaluation of chatgpt: https://t.co/fvZq4vM0Tn Basically, ChatGPT needs more work to help with urgent public health resources like suicide prevention and addiction.