Chat Hallucinations is a satirical site that lovingly celebrates and lampoons the current state of artificial intelligence.

Large language models are incredibly powerful.

But they are still in their infancy and are far from infallible. In particular, they hallucinate. That is, they generate plausible-sounding but entirely fabricated information. Or, as we like to put it, they present complete nonsense as fact.

AI models also demonstrate other quirks such as ignoring instructions, presenting answers based on assumptions, pre-empting questions by responding too soon, arguing against definite facts and engaging in unnecessary flattery.

"...as we humans unquestioningly embrace AI and hurtle towards our own self-induced obsolesence in a future where thinking has been forgotten and our minds are mush, we might as well have a laugh on the way."

All these quirks will no doubt be ironed out over time. Until then, as we humans unquestioningly embrace AI and hurtle towards our own self-induced obsolesence in a future where thinking has been forgotten and our minds are mush, we might as well have a laugh on the way.

So Chat Hallucinations presents a selection of AI-style interactions that highlight the foibles of our future digital overlords. All chats in Chat Hallucinations are entirely made up. They are satircal pieces written solely for entertainment.

Our Mission

Don't get us wrong. We love AI!

Satirising its quirks does not mean we don't like it. We do. We love it. We use it. It's great!

But highlighting AI's shortcomings is a useful reminder that we should never stop thinking for ourselves.

At Chat Hallucinations, we aim to:

• Encourage critical thinking when relying on AI-generated content.
• Highlight the current the limitations of AI systems.
• Foster discussion about AI safety and reliability.
• Write absurd, fictitious AI chats that should make you laugh.

Disclaimer

All AI chats on this site are entirely made up. Do not rely on any information set out in the chats in this site. They are 100% unreliable. Which is just a little less reliable than information from real AI sites.

Question 1

Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text.

Question 2

Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text. Answer text.