USC expert, S.C. residents push back on AI doomsday fears - WRDW

COLUMBIA, S.C. (WRDW/WAGT) - While national headlines warn artificial intelligence could pose an existential threat to humanity, South Carolina residents and students say their everyday concerns center more on privacy, loss of creativity, and digital dependency.

Recent broadcast interviews featuring computer scientists predicting that advanced AI could trigger catastrophic consequences over the next half-century have fueled widespread public anxiety. Yet on the ground in Columbia, people interacting with AI chatbots and algorithms daily say the real friction is already happening on their screens.

For college students navigating classes, AI tools have quickly transitioned from novelty to regular study aids. However, students say reliance on automated assistance can quickly backfire if not managed carefully.

“I think it can become detrimental when it’s used to generate its own ideas and you don’t have any of your own input in it,” said Kenzie, a University of South Carolina student. “As long as we include our original voice and opinion in it, then it won’t be detrimental. But as soon as it crosses that line and it becomes only AI taking over, that’ll be the problem.”

Fellow USC student Rory echoed that sentiment, noting how quickly social media platforms and search engines adapt to individual habits.

“What you interact with is how your algorithms and these social media platforms, even like Google, appears to you,” Rory said, pointing to targeted advertisements that surface immediately after casual searches or conversations.

Local entrepreneurs also find themselves balancing efficiency with unease. Sabreen Akram, a Columbia business owner, uses AI tools for marketing and operational questions, but worries about rapid advancements outstripping human control.

“Artificial intelligence is getting way over-intelligent, and I feel like it can exceed human intelligence,” Akram said. “If it can do that, what’s to say it won’t take us over or do something to take us out? It worries me.”

Other residents see a more immediate threat: the gradual decline of independent problem-solving. Dayon Sheehan, a Columbia resident who uses AI to brainstorm business phrasing and educational prompts for her children, fears younger generations may abandon foundational research skills.

“Humanity is just going to get more lazy in a way, and not just try to be creative and just give it all to the computer,” Sheehan said. “We still had the encyclopedia. We still had to research things.”

At the University of South Carolina, computer science professor Dr. Biplav Srivastava says the apocalyptic rhetoric coming from national talking heads obscures the genuine, manageable risks of modern software.

“To think that people have started building systems which will kill each other, and they will do all kinds of things, that is just fear-mongering,” said Srivastava, a professor at USC’s Artificial Intelligence Institute who leads the AI for Society research group.

Srivastava noted that society has adapted to transformative technologies before—from X-rays and automobiles to aviation and nuclear power—by implementing regulations, licensing, and consumer guardrails. Rather than autonomous doomsday machines, he described AI as an “accelerator of human intent.”

“If someone wants to hurt someone, they can find ways irrespective of technology. If I want to do good, I am not dependent on technology,” Srivastava said. “If you are just going to delegate everything to technology and somehow hope that magically it will figure out the right thing to do, when you yourself have not figured it out, God bless you.”

Instead of fixating on distant science-fiction outcomes, Srivastava advises consumers to practice baseline data hygiene. He cautions that commercial chatbots and digital platforms are driven by financial incentives to track user habits.

“Anything in AI starts with data,” Srivastava said. “For getting a one-dollar discount, if you are giving your email and your phone number to anyone you can see, then, of course, you are not following good data practices. The risks start from the data.”

He also compared artificial intelligence to cruise control in a vehicle: while the system can maintain speed, the human driver remains legally and practically responsible for where the car goes.

To help bridge the gap between complex computer science and the general public, USC’s AI Institute is expanding community outreach programs. The university will host an educational workshop on campus Oct. 23 specifically tailored to help seniors navigate AI tools, digital literacy, and online safety practices.

Copyright 2026 WRDW/WAGT. All rights reserved.