acm-header
Sign In

Communications of the ACM

ACM Careers

Grad Student Used a Neural Network to Write His Papers


View as: Print Mobile App Share:
documentation on circuit board background, illustration

A student working on a master's degree in business used GPT-2, the text-generating algorithm created by research lab OpenAI, to write essays for two different courses. He passed both classes.

He also used a tool called Talk to Transformer, which makes GPT-2 accessible through a web browser. "That's what made it possible for someone with a non-technical background like me to use GPT-2 to write my essay," says the student, identified by his first name, Tiago.

"It did get an okay grade," he says of GPT-2. "I passed and some students didn't. It was okay for me because I'm at the end of my masters. For me, it was just pass or fail."

Tiago wrote an outline of his essay and the first sentence of each paragraph. He fed the sentences to GPT-2 which returned a full paragraph.

"What I figured as I was doing it was that if I write my first sentence in a certain way, it increases the probability that the paragraph will look how I want it to," he says. "I found adding words like "innovation," "synergy," and stuff like that made the essay sound more suited for the course."

"The final essay is pretty poor, it's just not poor enough for the professor to fail you," he says.

From Futurism
View Full Article


 

No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account