June 01, 2023 • By The UCI Paul Merage School of Business
The benefits and drawbacks of ChatGPT and other forms of artificial intelligence were recently discussed during the inaugural Digital IQ Series event.
The Center for Digital Transformation (CDT) launched the seminar series to demystify and explain exciting new technologies to members of the Merage School and the broader business community. Each event will feature a knowledgeable expert who will discuss the promise and risks of emerging technologies.
“We have had a long-standing demand from our constituents that they wanted to understand more about emerging technologies and how they impact society and so we decided to launch this brand new series,” said Vijay Gurbaxani, Director of CDT. “We are calling it the Digital IQ Series for obvious reasons, we are trying to raise our digital intelligence and provide a forum for the community to think about and discuss these issues.”
During the first event, Merage School Professor Vibs Abhishek spoke about Generative AI, in particular focusing on the viral sensation ChatGPT. Generative AI produces text, images, and other media based on prompts.
“With Generative AI, we are at a place where technology feels almost magical,” Abhishek said. “It has a lot of implications not only for what it can do but also the impact it will have on society.”
Abhishek said that ChatGPT can answer questions, write poetry, generate cover letters, and even play a murder mystery game. It’s also differentiated from its predecessors by how long somebody can have a conversation with it.
Abhishek said that Generative AI programs like ChatGPT create written works that are better than the work of people in a lot of instances. He said that AI programs are now capable of passing the Turing test, which assesses whether a computer is distinguishable from a human.
“If the computer is able to fool you into thinking it's a human being, it has passed the Turing test,” he said. “I think we have passed that very safely. So when you're talking to these things, like ChatGPT, you don't really distinguish that you're not talking to a real person. That's how good these systems have become.”
However, imagery and computer coding produced by AI still needs to evolve. Another outstanding issue with the program is it doesn’t guarantee the accuracy of its answers.
“There needs to be a layer built in which makes sure that the answers it gives out are actually correct,” he said.
Abhishek said that AI has been largely used in the work world for repetitive tasks but Generative AI is enabling more creative applications of the technology.
Generative AI can help many professionals with their work. For instance, a copywriter could use it to improve marketing copy or a graphic artist could use a program to create an initial image to guide their work.
Within health care, Generative AI can help in a number of ways, including medical education assistance, triage support, patient follow-ups, mental health support, administrative tasks, and telehealth services.
“I think in the very near future, Generative AI will be able to solve a lot of tasks that other forms of AI have not been able to solve,” he said. “We have seen that it leads to much more efficient help, so people can find help much quicker. It also leads to a dramatic reduction in the support team companies need to hire.”
While Abhishek believes that Generative AI has a lot to offer, it also may jeopardize a variety of jobs, including customer service representatives, content writers, translators, editors, social media managers, telemarketers, travel agents, accountants, graphic designers, paralegals, sales representatives, and data entry clerks, among other occupations.
“There are many professions that I feel are going to be threatened by some of these advancements,” he said. “I think it’s not going to be that AI will replace all jobs in these fields but only a few people will be needed to do these jobs that a larger number of people were doing in the past. I think that’s a big cause for concern.”
Abhishek said another issue with the program is that it’s not available in many countries, so people who have access to Generative AI will be able to do very different types of things and will have very different outcomes than people who do not have access. He believes this will increase the digital divide.
Abhishek also said that the AI programs could perpetuate discrimination because they are trained on Internet data that is inherently biased. If leaders make decisions based on the models, then the bias could be amplified in society.
“It’s really fascinating what Generative AI can do but as a society we should think very carefully about how to harness the power of this type of technology and how to make it accessible to everyone,” he said. “If we don’t make it accessible to everyone, we’re just going to create a very bifurcated world.”
Associate Director of Communications
jrotheku@uci.edu