What has Google done that is so great or dangerous with GPT-3?

Its like having access to an infinite army of 15 year olds who can read and write pretty decently but have no clue how ignorant they are. But but its infinite. So what? Take a bunch of ignorant 15 year olds and get some actual work done in the real world and keep tabs on how much energy it takes.

People are full of shit when it comes to threat perception these days. Default position is anything and everything new and not understood is a threat. Sort of like watching chimps in a tree jabbering about shadows.

As for this notion of catastrophic disinformation. It doesn't matter how much disinformation or propaganda one chimp does if the entire chimp troupe is also busy pumping out disinformation and propaganda. It all cancels each other out - https://ubiquity.acm.org/article.cfm?id=3061712

This is a far better description of the underlying project: https://en.pingwest.com/a/8693, "Chinese AI lab challenges Google, OpenAI with a model of 1.75 trillion parameters"

Fair's fair. The most cost effective solution is to hack Chinese servers and practice intellectual property theft. You know, pick the flowers and bring the honey back. Or perhaps some kind of 1000 talents program where some Chinese AI researchers only pretend to work for China.

One of these days this race to infinity is gonna end badly, I can sense it.