ChatGPT Error Could Have Cost a Life

In the US, a blogger shared a story that sends chills down your spine: her friend almost suffered due to ChatGPT's advice. The girl sent the bot a photo of a plant found in the yard and asked if it was dangerous. The AI confidently stated that it was harmless carrot leaves — and repeatedly denied that it could be poisonous water hemlock

The problem is that water hemlock is a deadly poisonous plant with no antidote. Fortunately, the friend decided to double-check the information: Google Lens immediately indicated that the plant is poisonous. Moreover, in another chat, ChatGPT later admitted its danger.

The story ended well only because the person did not fully trust the AI

What do you think?