[strobe999]

Having read several articles and comment threads about the ChatGPT research preview, the general trend seems to criticize its accuracy. I don't see how it even matters.

Traditional internet search engines are biased towards the truth. Search for lies on any given topic and you will find a page telling you about the real truth. Rarely an obvious disclaimer about truthfulness or accuracy of the content.

This language model on the other hand seems doesn't even claim to know the truth. Without asking, it will go out of its way to tell you that it might be wrong for an arsenal of reasons.

With such an ever-present caveat, humility, ChatGPT is what I would call human friendly and what google & co should've been. A search engine acting in the inquirers best interests. Not only sharing knowledge but most importantly - perspective.

I am personally looking forward to a future where this and similar models are readily available to non-technocrats. I imagine using this instead of google for general knowledge would reduce radicalization on any axis and generally broaden empathy.