West Virginia Legislature

September interims: Panelists lead legislators in conversation on artificial intelligence

MORGANTOWN – A panel of experts led legislators on an exploration of the evolving world of artificial intelligence during Monday’s interims.

House Speaker Roger Hanshaw opened the joint House-Senate general meeting.

“If anyone is under the impression that artificial intelligence and generative AI hasn’t already changed our world, you’re living in a fantasy land,” he said.

(Generative AI can produce content such as text, images and audio – along the lines of ChatGPT.)

Moderator Brad Smith – Marshall University president – cited the military acronym VUCA, saying we live in a volatile, uncertain, complex, ambiguous world and thriving that requires agility and the willingness to experiment. “There’s never been a more important, game-changing technology,” he said. “It’s touching every corner of the society.”

The four panelists described what they see as some of the most promising current AI applications.

Jamie Butler, with Amazon, said AI can help government be more effective: fight fraud and enhance citizen services. California’s DMV, for instance, was able to improve its performance by using AI to make the licensing experience self-service.

Suj Perepa, with IBM, talked about the growing prevalence of virtual agents across all industries. They can emulate human-to-human interactions and leave humans to do more high-value work.

Ryan Palmer, with Microsoft, said AI can also change lives on a very individual basis. It can make life more accessible for people with disabilities – for instance a phone app for the visually impaired that can take a phone picture and give the person an instant verbal description of their surroundings.

Amy Cyphert, with WVU law school, said AI can help create a more just and equitable society, but AI depends on the data fed into it and could also exacerbate bias. “It’s all going to come down to how we choose to use it, and a lot of that will depend on how you chose to regulate it.”

The conversation moved to implications for the workforce. Perepa said that in 2020 the World Economic Forum predicted that by 2025 AI would displace 85 million manual jobs but create 95 million higher-skill jobs.

Employers and employees will need to embrace the value AI can bring, she said, and re-strategize workers’ skillsets and roles. “AI is not here to manage us.” AI is here to augment human intelligence and efficiencies, never to replace humans, she said.

AI can help relieve works of menial, manual tasks. Digital labor will involve humans and AI working together. We will be able to create new drugs faster. By 2030, one of every 10 cars will be autonomous.

But AI growth must be guided by what she calls ERT: It must be ethical, responsible and trustworthy.

Cyphert also cautioned against over-reliance. “Creation is a human endeavor and creativity is a human characteristic. … These tools should support that, not supplant that. … If we overly rely on these tools, we lose some of that innate human creativity.”

All four agreed AI has and must continue to have its limits. “AI cannot make inferences or decisions,” Perepa said.

Returning to ethics, Palmer said the question must not be what AI can do but what AI should do.

Cyphert talked about AI’s potential for invasion of privacy, noting our laws tend to be insufficient and patchwork. Apps and websites, for instance, often require us to agree to terms and conditions, and we don’t read those, or know what kind of data those apps and sites are taking from us.

A better solution for the future would be to opt-in to data gathering rather than opt-out, she said. That way we retain some control and have some idea of what we’re giving away.

Email: dbeard@dominionpost.com