For the past seven years, software company Problem Solutions has been conducting funded research on AI applications to improve companies’ performance and bolster human capital, company founder Mike Hruska said.

His Johnstown, Pa., company builds proprietary artificial intelligence applications for the federal government and the private sector that are more fine-tuned than the public, open and free AI, including ChatGPT, Hruska said.

ChatGPT is pre-trained with information through September 2021, so it can’t answer, for example, “What was the weather like in Corbin, Kentucky, on Oct. 1, 2021?”

Developers can purchase access to a ChatGPT model and further fine-tune it with real-time information such as weather forecasts.

Developers can also train the program to behave in certain ways, for example, with the personalities and tendencies of Mr. T, Samuel L. Jackson, David Goggins or a Marine Corps instructor.

That’s exactly what Problem Solutions has done with Mr.GPT, a bot specifically trained to give users harsh, motivational advice to help people reach goals.

Mr.GPT is a public example of what the company is working on privately, Hruska said.

“We teach bots to talk about some things and not talk about other things,” he said. “We teach bots to have attitudes and behaviors and I say bots that are largely text-based, but we are doing this with meta-humans and holographic technology that really simulates real life.”

Judging the competition within his own industry, Hruska said he’s not worried about tech giants reducing innovation by cornering the market.

“I’m not scared of large, slow organizations, Hruska said. “We build stuff with two- or three-person teams in days or weeks that large teams can’t build in even four to six months. The agility of small, high-performance teams is key.”

Hruska is working with several authors to build AI models that know all of their books and can help write new books in the same style, he said.

“AI is no longer a technological problem,” he said. “It’s merely a matter of engineering.”

With the pace at which AI is progressing, Hruska foresees “meta-humans” replacing humans in call centers in three to five years.

“Meta-humans are in development which talk to you in the way you talk to it, it mirrors you,” Hruska said.

‘Take over the world’?

Concentrix is the world’s second-largest provider of customer experience solutions. Its 2022 revenue was $6.3 billion.

When speaking at the June Bank of America Global Technology Conference about the impact of generative artificial intelligence on customer experience industry, Chris Caldwell, president and CEO of Concentrix, said clients big and small are talking about AI.

However, few companies are jumping in with both feet.

“There’s quite an amount of trepidation around how to use it, the security around it and how to get predictable results,” Caldwell said. “So we don’t have clients out there now who want to blaze the trail and say this will take over the world some day.”

One thousand companies, including 130 fortune Global 500 companies – the largest companies in the world by revenue – outsource customer experience-related tasks to Concentrix.

Based in California, Concentrix has more than 300,000 staff members worldwide.

Concentrix’ clients who say AI can replace humans are in the minority, Caldwell said.

“There’s not one of the enterprise brands that’s willing to risk their reputation by showing up on the front page of the paper by giving some result that is either not aligned to the brand or against social consciousness or anything else that goes along with that,” he said.

The company’s services include collecting and analyzing customer feedback to improve their clients’ customer experiences and overall business performance.

Concentrix automated 10 percent of its contract volume last year through AI machine learning.

Generative AI is the next evolution, Caldwell said, but necessary infrastructure “hooks” to enable it are costly.

“For some very basic tasks, generative AI costs more than an off-shore worker (aided by machine learning),” Caldwell said. “So really, it’s not something that’s suddenly going to take out a lot of low value work as we would have liked.

“Where we see most of our client conversations about AI is really around how to make staff more productive and proficient, faster,” he said.

Josh Cragle has been in customer service for roughly nine years. He currently works in health insurance customer service from his home, which bustles with activities of a puppy and his children.

He’s not fretting over the cloud of meta-humans that could be approaching in the distance.

“We really haven’t heard much about AI taking completely over in the future,” he said. “As long as health insurance is around, you are still going to have to talk to people on the phone to get your answers.

“I’m not too concerned about being replaced by a bot.”