Curiously Intelligent

A recent article in the Süddeutsche Zeitung by Sarah Spiekerman has caught the attention of the masses. Diederichs-Verlag has released a book that will get a lot of attention; it was mostly authored by an artificial intelligence, while humans also contributed as co-authors.Iain Thomas and Jasmine Wang edited the volume entitled What Makes You Human. In it, the unnamed Artificial being provides startlingly insightful responses to some of humanity's most important concerns, like "Why do humans suffer?" "How do I make peace with my body?" and "How do I know if I'm successful?" Yet in the case of AI, is it really appropriate to use the term "being"? Or does one attribute a nonexistent being to the computer that is using the GPT-3 language model?

Let's start out by being crystal clear: The responses provided by the AI, which Thomas and Wang trained using particular wisdom texts, are indeed astonishingly brilliant. clever enough to wow anyone who is at least moderately educated. For instance, the AI responds to questions from human authors such as "Why do we suffer?" by saying "Suffering arises from the illusion that we are separate from the universe", and "How does love grow?" by saying "By giving". The AI will occasionally provide extremely extensive replies that span a full paragraph or page. For instance, in response to the query, "Can man be good?" AI responds: "To be good is to be in harmony with the universe. Harmony is virtue, virtue is good, and good is the natural way of the universe. The natural way works.".

These responses are based on "chosen extracts" from classic works of literature like the Bible, the Torah, or the Koran.

Read more here.

‍