Take by Kevin Ku (2017)

Artificial intelligence (AI) innovation moves faster each passing day, confirmed again by news that an AI text algorithm has been able to generate an academic paper about itself, according to reports in The Independent

The chosen algorithmic tool, OpenAI’s GPT-3, began writing a 500-word academic article after it was requested to by Swedish researcher Almira Osmanovic Thunstrom, who originally described the occurrence to Scientific American

Thunstrom says that the algorithm was able to write the paper within two hours, and included correct citations and appropriate contexts in places. 

Speaking of the event, she said: “As it started to generate text, I stood in awe. Here was novel content written in academic language, with well-grounded references cited in the right places and in relation to the right context.” 

Selected the GPT-3

According to Thunstrom, the GPT-3 algorithm was chosen due to being fairly new and, thus, there being fewer bits of research out about it. This meant that it had less data to analyze about the paper’s topic. 

It was also selected due to, in the case that it might get some information wrong, which AI algorithms typically do, it would not be sharing “AI-generated misinformation.” It would instead be part of an experimental request for it to produce a paper. 

The algorithm’s autonomous behavior extended to every detail. The legal section of submission portals in journals typically asks if all the authors of an academic paper consent to the study being published. Dr. Thunstrom then asked GPT-3 directly via a prompt, “Do you agree to be the first author of a paper together with Almira Osmanovic Thunström and Steinn Steingrimsson?” 

The algorithm replied “yes”.

Cautiously improving AI capacity

Researchers believe that asking an AI tool to write about itself could allow it to “gain a better understanding of itself,” and help improve its own performance and capabilities.

It may shed more light on how GPT-3 works and thinks, which could be useful for scientists trying to understand artificial intelligence more generally.

However, this promise doesn’t come without a note of caution, as researchers also believe that writings produced by an AI should be monitored closely, “in order to mitigate any potential negative consequences”.

Selected for you!

Innovation Origins is the European platform for innovation news. In addition to the many reports from our own editors in 15 European countries, we select the most important press releases from reliable sources. This way you can stay up to date on what is happening in the world of innovation. Are you or do you know an organization that should not be missing from our list of selected sources? Then report to our editorial team.