The AI Mirror Test

Introducing the AI Mirror Test, which very smart people keep failing – Great article from The Verge on AI and our tendency to anthropomorphize machines that mimic human behavior, even when we’re fully aware we’re talking to a machine, also known as the ELIZA effect.

AI chatbots like Bing and ChatGPT are entrancing users, but they’re just autocomplete systems trained on our own stories about superintelligent AI. That makes them software — not sentient.

A monkey looking into a mirror at its own reflection

Who’s that in the mirror?

Just like humans, these new AI chatbots at times seem very intelligent, and sometimes just outright stupid – but we should not forget these are just algorithmic responses and not a sign of any sentience. We are still a very long way from that point – if it ever will happen.

Source: Introducing the AI Mirror Test, which very smart people keep failing