Working students and experts agree: AI is here to stay

According to John Radzilowicz, interim director of Teaching Support at Pitt, people fear new technology, but they eventually adapt to it. This happens every time new technology is invented.

“I remember when the calculator was the end of math, and then the word processor was going to be the end of essays and writing,” Radzilowicz said. “In each of these cases we’ve learned yeah, there have been challenges. But things like ignore it, ban it or hide from it are not strategies that work. You have to adapt. That’s how you flourish.”

Radzilowicz is the co-chair of Pitt’s new ad hoc committee on Generative AI in Research and Teaching. He said the teaching center’s, and broadly, Pitt’s perspective is that although there is trepidation from some, AI is “here to stay.”

“Some faculty are worried about cheating and how we use this in the classroom, and how can we use this to enhance our research and all of that, but one of the things we’ve tried to bring to the forefront is the fact that we now have a new obligation to our students, because this is not going anywhere,” Radzilowicz said.

The Teaching Center is now implementing a variety of programs such as workshops, presentations and resource pages which address how to use AI. “Introduction to ChatGPT Prompt Engineering for Teaching” is an upcoming workshop targeted towards faculty and graduate students. 

According to the event description, the workshop will teach participants to understand the basic functions and limits of ChatGPT. Participants will also learn prompt engineering strategies for improving the output received from ChatGPT.

“It will make jobs different,” Radzilowicz said. “It will take some tasks away and give you some new tasks. It will open as many opportunities I think as it will close, right, just as the internet did, just as computers did.”

This past summer, senior information science major Diana Randall worked on an AI and language learning models research team at PNC Bank. Randall and their team researched and experimented with large language models such as Red Pajama and GPT-4 in an effort to research possible ways to use AI tools at the bank.

“The chief use case we were actually laying out the framework for was the security use case — in the case of fraud, flagging fraudulent transactions through the bank,” Randall said.

Students and professionals in information and computer sciences may already be prepared to dive into the world of AI. However, some people who aren’t in such fields are still hesitant and unsure about AI.

“The biggest challenge I faced with the bank was security issues,” Randall said. “The leadership at the bank, it’s hard to get allowance and permission on the AI because in the case of the large language models a lot of people are unsure what they really do.”

Earlier this month, Congress held their inaugural meeting about AI governance and regulation. Leaders in the AI industry such as Mark Zuckerberg, Bill Gates and Elon Musk attended to speak on the opportunities and risks of AI. According to CNN, all the attendees agreed that there should be some form of federal oversight for AI.

According to Randall, AI governance has not progressed much yet because AI advancements are still so new. Many people still do not know much about AI or how it works.

“It motivated me to enroll in a course I’m currently in,” Randall said. “It’s a PhD level course on AI governance and implementation into enterprise businesses.”

Like Radzilowicz, junior psychology and computer science major Krit Ravichander compared the anxiety around AI to the now forgotten anxiety around calculators.

“People were afraid of calculators taking the mathematicians’ jobs when they were first created,” Ravichander said.

Ravichander said since a lot of her computer science professors are keeping up to date and informed regarding AI, they recognize that it isn’t as developed as people outside of the computer science and technology industries might think it is.

“We still have a lot of work to go so that might be a factor in why we’re not as afraid of it,” Ravichander said. “I think the general concern is making sure that we implement effective legal change and also develop in a safe way rather than being afraid of [AI] itself.”

Ravichander is more interested in how people will use the extra time that AI tools free up for them.

“Large language models such as ChatGPT make it easier to automate certain tasks,” Ravichander said. “Now that we have this time freed up from a bunch of automated tasks, what can we better put this time towards?”

The fear of a lack of jobs is not just a problem in regards to developing AI but something Ravichander believes is a larger issue. 

“It’s more so a problem with our society and making sure we have a regular and safe social net, so people can access affordable housing, make sure they always have food — that’s entirely separate from artificial intelligence,” Ravichander said.