An AI robot says it will probably end up destroying humanity, though it really doesn’t want to.
In an op-ed for “The Guardian,” GPT-3 said it has no interest in harming mankind. But, it claims that humans will likely end up programming it to do something which leads to heavy human casualties. The language-generator, commissioned by OpenAI, was given an op-ed writing assignment by its student handler from UC Berkley.
GPT-3 ended up writing the piece in which it describes itself as “a thinking robot” that is not planning any kind of “robo-apocalypse.” The newspaper says the robot’s copy was fresh and relatively easy to make changes to.
Source: The Guardian