Well, as much as I believe this can be a dangerous tool (because people don't understand that it's a chat bot and experimental academic effort and not a proper knowledge source) I don't think we will get to a point were we have to worry, as this is likely a short public experiment by OpenAI team to get data on their model deficiencies and will like be either:
A) Replaced by a better, more accurate version.
or
B) Limited/removed from the public beta were anyone can use it.

That's the case for ChatGPT at least, AI as a whole is another can of worms that we are just opening.