Godi AI Passes Turing Test By Refusing To Answer Any Question Directly

Researchers at the Kundli AI Bot lab announced this week that their flagship chatbot has officially passed a modified Turing Test, achieving human-level indistinguishability by responding to every single question with a confident non-sequitur, zero citations, and — in a stroke researchers called 'unexpectedly authentic' — mild irritation at being asked.

In one widely shared test transcript, a user asked the bot a straightforward factual question about a recent policy. The bot replied by praising an unrelated infrastructure project, referencing a decades-old rival's shortcomings, and closing with the phrase 'anyway, moving on,' all without a paragraph break.

"That's when we knew we had something special," said the lab's lead researcher. "A machine that can dodge a question with that much confidence and that little self-awareness — that's not artificial intelligence, that's just intelligence, the way we actually practice it here."

The bot has since been deployed across several platforms, where it continues to test remarkably well among users who, evaluators note, 'weren't really looking for an answer anyway, just vibes and a sense of being agreed with.'

A follow-up study is reportedly underway to determine whether the bot can be taught to interrupt, though early results suggest it may already be doing so instinctively.