Published On: Mon, Mar 25th, 2019

Robot Authors Not a Threat Yet

Totally machine-generated prose came a step closer to reality with the announcement of an artificial intelligence that could produce news stories or fiction — but writers and reporters don’t have much to fear just yet, author Steven Poole argues.

In a commentary for the Guardian, Poole chided OpenAI, a nonprofit lab backed by Elon Musk and other tech entrepreneurs that just unveiled its AI called GPT2 —  for suggesting it was too dangerous to release lest it created “deepfakes for text.”

“Let’s first take a step back,” Poole argued. “GPT2 is just using methods of statistical analysis, trained on huge amounts of human-written text…to predict what ought to come next. This probabilistic approach is how Google Translate works, and also the method behind Gmail’s automatic replies.”

“It can be eerily good, but it is not as intelligent as, say, a bee,” Poole wrote. “Right now, novelists don’t seem to have much to fear.”

As a news reporter, GPT2 is “rather Trumpian,” Poole added — garbling context that its creators admitted is a “world-modelling failures… the model sometimes writes about fires happening underwater.”

“The excessive hype surrounding the GPT2 text generator, at least, is a symptom that we have to some degree internalized the philistine functionalism of Silicon Valley, according to which everything is simply data,” Poole argued. 

“But writing is not data. It is a means of expression, which implies that you have something to express. A non-sentient computer program has nothing to express.”

Let’s block ads! (Why?)

Newsmax – Science and Technology

Leave a comment

Robot Authors Not a Threat Yet