Chatbot Recommending Poisonous Recipes

3 pointsreggie_eu1 comment

Have there been any product liability suits for chatbots recommending poisonous recipes? I was able to get both Bard and ChatGPT to recommend some toxic lasagna.

View on HN

Can you provide examples? if you ask explicitly some toxic ingredients and stuff.... yeah I doubt we can consider the "AI" to be in direct fault.

If a kid asks "how do I bake cookies for my family" and the AI proposes to add some cyanide into the mix... yeah we have an issue.

Reply on news.ycombinator.com