AI Experts lay out Tech’s Common Problems

Artificial intelligence (AI) is coming full force into the technology world, especially since it allows the workforce to leverage the emerging technology to focus on higher-level tasks by using tools like automation and machine learning. But what exactly is AI and what are some of its common problems?

During Dell Technologies World 2023, two AI experts sat down to discuss the most common problems they see within emerging technology when they are working closely with customers.

First and foremost, Marc O’Regan – Dell’s chief technology officer (CTO) for Europe, the Middle East, and Africa – said that clearly defining AI is the most common problem he sees amongst his customers.

“We’re hearing so much about artificial intelligence, depending on who you talk to, whose hands you go into, and what couch you sit down on,” O’Regan said during day two of the conference in Las Vegas.

The CTO explained that AI is the designing and building of intelligent agents that receive percepts from the environment and take actions that affect that environment, and within that umbrella falls machine learning and deep learning – both subsets of AI.

“AI is built ultimately to solve complex problems – they could be business problems, organizational problems, industrial problems, or even societal problems,” O’Regan explained. “Or create something nascent and new, something that we haven’t been able to think about or haven’t been able to achieve before.”

While the AI experts said that defining the tool is the most common problem, other use case issues they noted with the emerging technology include: confusion about architecture; sharing infrastructure; teaming and tooling; resource and knowledge management; silos; and generative AI.

When the experts polled the audience, the most common problem was data prep, quality, and organization.

“A lot of people have the notion of more data is better, but more data is not necessarily better, especially if it’s the wrong data,” said Mike King, an AI solutions architect for Dell. “Applying data quality to the process can improve your throughput and raise your success rate for an AI project.”

“I couldn’t agree more,” O’Regan said. “Getting good quality data ready and prepared and using technology to be able to do that and automate some of that process.”

“If you talk to data scientists, some of them will tell you they’re spending up to nearly 80 percent of their time just trying to wrestle with the data – clean and wrangle it down, prepare and do the data science, do the mathematics,” O’Regan continued, adding, “Twenty percent of their time is doing what they really should be doing. But what they should be doing is trying to solve problems within the organization within the business.”