Open AI made headlines when they announced that AI was too dangerous to release in the wild. AI is a potential intelligence threat and as far as novelists are concerned, they may just put the writer out of business. How does it work?
GPT2 is just using methods of statistical analysis, trained on huge amounts of human-written text – 40GB of web pages, in this case, that received recommendations from Reddit readers – to predict what ought to come next. This probabilistic approach is how Google Translate works, and also the method behind Gmail’s automatic replies (“OK.” “See you then.” “That’s fine!”)
AI doesn’t seem to have an Orwellian or Austenian touch. When applied to predict sentences that follow landmark lines, the prose that has been generated is dry and ridiculous. Poole doesn’t feel threatened enough by AI-induced potentially disastrous fake news as misleading articles are already the norm. It seems to be more a case of not having to pay a royalty to bot writers as opposed to writers in the flesh than any serious bot threat posed by them yet.
Read The rise of robot authors: is the writing on the wall for human novelists? by Steven Poole.