Half of #AI #health answers are wrong even though they sound convincing#Chatbots, #ChatGPT, #Gemini, #Grok, #MetaAI and ...

Half of #AI #health answers are wrong even though they sound convincing#Chatbots, #ChatGPT, #Gemini, #Grok, #MetaAI and #DeepSeek, asked 50 health and #medical questions spanning cancer, vaccines, stem cells, and nutrition. Two experts rated every answer. Nearly 20% were highly problematic, half were problematic, and 30% were somewhat problematic. No chatbot reliably produced accurate references, and only 2 out of 250 questions refuse to answerhttps://theconversation.com/half-of-ai-health-answers-are-wrong-even-though-they-sound-convincing-new-study-280512https://archive.ph/UDGvi

Read Original

Related

Mastodon discussion 10m ago

この大事な時期に日本に巻き込まれて被弾でもしたら……Apple、試合のスコアや成績などの情報をリアルタイムで表示してくれるスポーツ情報アプリ「Apple Sports」に日本のJ1リーグの情報を追加。 https://applech2.co...

この大事な時期に日本に巻き込まれて被弾でもしたら……Apple、試合のスコアや成績などの情報をリアルタイムで表示してくれるスポーツ情報アプリ「Apple Sports」に日本のJ1リーグの情報を追加。 https://applech2.com/archives/20260721-apple-sports-support-j1.html#Apple #LLM ...