The United Nations’ High Commissioner for Human Rights Michelle Bachelet is lobbying for a temporary suspension on usage and sale of Artificial Intelligence systems, according to National Public Radio. BGSU’s campus is no stranger to this branch of computer science, as it is seeing integration of new artificial intelligence products in the form of Starship Robots, and may very well be home to more advanced tech in the coming years.

Bachelet’s request comes in response to a United Nations report out of Geneva detailing unaddressed risks of AI.

The field of AI was formally founded at Dartmouth College in 1956. It is commonly defined as the ability of machines to self-learn, without explicitly being programmed to do so. In the years since, AI and its sub-category, machine learning, has been progressively weaving their way into the fabric of everyday life. They manifest through such functions as the recommendation systems for YouTube and Amazon.

BGSU is currently utilizing AI in the Electrical Engineering department of the College of Architecture, Construction Management and Engineering through industrial-style robotics. The College of Education also has a form of AI to facilitate the field experience for freshmen.

The most recent implementation is Starship Robotics. The fleet of food delivery bots was launched on campus on April 22, 2020. According to John Ellinger, BGSU’s Chief Information Officer, the choice to implement the service was an easy one.

“I was in the room 21 months ago, when we had our first discussion about, should we or shouldn’t we bring (Starship) on-campus? I was absolutely for it, simply because it is a working practice of artificial intelligence that is a daily reminder, or visible, to students,” he said.

The bots utilize artificial intelligence systems in their navigation, constantly identifying and adapting to the blockages and obstacles in their path at any given moment.”

What one robot learns is then communicated to the entire Starship fleet.

“When that robot senses the blocking, it records it, sends it back to the mothership and the mothership then sends it to all the other Starships saying that ‘if you are in this GPS, this location, this walkway or street, is blocked and you need to go around it,” he said.

Starship’s delivery locations have since expanded to reach beyond campus, with outer boundaries nearby the Wood County Regional airport, Falcon’s Pointe Apartments, I-75 and Haskins Road.

However, some students did have doubts. Fiona Warner, president of BGSU’s chapter of Women in Computing, expressed some uncertainty regarding the “implicit data processing.”

“The robots sparked some debate within the tech community, because students didn’t really get to opt in to Starship’s use of the robots on-campus. Someone was saying how when we are becoming a new student, there’s probably some line in some contract that we had to sign where we gave up the right to control whether our data is used by third-party services on campus,” she said.

Despite the worries, Warner sees potential for AI to bring more ease into the lives of BGSU students, aside from food delivery.

“If I had an AI that could help me plan my schedule, that would be great. I wouldn’t have to spend hours looking at my degree audit, trying to figure out what classes I have to take,” Warner said.

Integrating more AI would present the campus community with heightened convenience. Others say it may also present unique challenges and inherent risk factors. Shuteng Niu, an associate professor in the Department of Computer Science, classifies these risks into three categories.

“The first is interpretability. The second is ethical issues and the third is insufficient data. For most artificial neural networks, they are more like black boxes. We don’t really know what is going on in there. So, will you trust enough to apply a black box to some critical applications?” he said.

Deferring to the judgment of faculty, their research and business models will likely determine where AI manifests on campus next.

“If we can fully understand how those machines [work], how those algorithms work inside, they are 100 times more stable, more reliable, than human beings. Right now, the challenge is we don’t really understand how they work inside,” Niu said.