A team of researchers from OpenAI — a San Francisco based research institute backed by notable Silicon Valley figures like Elon musk and Peter Thiel — has created an artificially intelligent writing system called GPT-2 that is able to produce coherent articles on a wide variety of subjects, requiring only a brief prompt to do so. The system’s Artificial Intelligence (AI) is also “unsupervised”, meaning that it does not need to be restrained to produce content.
However, the researchers have opted to withhold the technology, saying it could be used for “malicious” purposes such as mass-producing fake news.
The system generates text using data taken from approximately 8 million webpages. To sustain the system’s information requirements, the OpenAI team created a new automated method to detect quality content on the Internet.
GPT-2 is able to disregard low quality content and only look at pages posted to the link-sharing site Reddit. From Reddit, it only uses links that have received three or more “upvotes” (the reddit equivalent of a Facebook like).
“This can be thought of as a heuristic indicator for whether other users found the link interesting, educational or just funny,” the research paper published by OpenAI said.
The AI then uses this data to generate articles or story word by word. This usually results in a cogent — but mostly false — body of text. This is because although the sentences it generates are based on information found online, the composition of its own articles is meant to be unique. Sometimes this results in passages of text that do not make a lot of structural sense or contain laughable inaccuracies.
The following are excerpts from a story written by the system:
The scientist named the population, after their distinctive horn, Ovid’s Unicorn. These four-horned, silver-white unicorns were previously unknown to science.
Now, after almost two centuries, the mystery of what sparked this odd phenomenon is finally solved.
Dr. Jorge Pérez, an evolutionary biologist from the University of La Paz, and several companions, were exploring the Andes Mountains when they found a small valley, with no other animals or humans. Pérez noticed that the valley had what appeared to be a natural fountain, surrounded by two peaks of rock and silver snow.
Pérez and the others then ventured further into the valley. “By the time we reached the top of one peak, the water looked blue, with some crystals on top,” said Pérez.
This text was generated using the following human-written prompt:
In a shocking finding, scientist discovered a herd of unicorns living in a remote, previously unexplored valley, in the Andes Mountains. Even more surprising to the researchers was the fact that the unicorns spoke perfect English.
“We have observed various failure modes,” the team observed. “Such as repetitive text, world modelling failures and unnatural topic switching.”
OpenAI has said that it wants its technology to prompt a debate about how AI writing systems can be used and abused, saying “We think governments should consider expanding or commencing initiatives to more systematically monitor the societal impact and diffusion of AI technologies, and to measure the progression in the capabilities of such systems.”