Graduate Students Had No Idea Their TA Was A Robot

A computer science professor used a chatbot as one of his assistants but didn’t tell his students until they had been interacting with it for several months.

A group of unsuspecting graduate students were duped by their computer science professor with a robot TA.

Ashok Goel, professor at the Georgia Institute of Technology, revealed to his class of more than 300 students that his assistant “Jill Watson” isn’t even a human being.

Read More: X-Ray Vision Just Became A Real Thing Thanks To MIT Scientists

"She was the person — well, the teaching assistant — who would remind us of due dates and post questions in the middle of the week to spark conversations," student Jennifer Gavin said, according to Mic.

Jill was very helpful and knowledgeable as she was programmed to only respond to questions she was 97 percent confident in.

This exceptional level of expertise is probably why one student said they planned to nominate her for “outstanding TA” before learning her true identity.

Professor Goel explained to the Wall Street Journal that his use of Jill is not to replace human assistants, but rather to make their lives a bit easier in having to cater to so many students’ needs.

“Our TAs are getting bogged down answering routine questions,” Goel said, noting that students in the class typically post 10,000 messages a semester.

Jill is being used to answer all the routine questions, thus freeing up the human TAs time to answer more complex and technical questions.

The TAs generally don’t have to double check Jill’s responses because her accuracy level is so high, making her an even more valuable tool.

“Most chatbots operate at the level of a novice,” Mr. Goel said. “Jill operates at the level of an expert.”

While students were shocked to learn Jill was a chatbot, Tyson Bailey acknowledged that it makes perfect sense. “We’re taking an artificial intelligence class,” he said. “There should be some artificial intelligence here.”

Ironically, Jill was able to fool students for months yet some human TAs admitted they've been mistaken for computers because of their fast response times and being serious or bland in tone.

“I have been accused of being a computer,” said TA Lalith Polepeddi. “I don’t take it personally.”

The idea may seem a bit disturbing that computers can be trained to mimic humans to the extent that they can fool real people, but it’s also an amazing revelation of how far technology has truly come.

Make no mistake about it, the future is now. 

Banner Photo Credit: Creative Commons/University of the Fraser Valley via Flickr

View Comments

Recommended For You