Kathy Gibson reports from Gitex – Artificial intelligence (AI) is the big buzzword of the day – but it will only be successful when and if organisations use it get value from their data.

There is a tremendous amount of noise about AI at the moment, but Joe Reis, founder and CEO of Ternary Data, thinks the companies making the least noise are often the ones already gaining value from AI.

“The underlying use cases for generative AI (GenAI) are there,” he tells delegates to the AI Anywhere conference at Gitex Global 2023. “But if you are still having a lot of steering committee meeting about your AI initiatives, you are probably not very far along.”

Polls show that about 40% of CEOs are talking about AI today – but this time last year they were talking about Web 3.0, Reis comments.

“We see that the GenAI space is still very immature. And those companies that are adopting it for the wrong reasons will face some real challenges.”

But, where organisations can find positive use cases, they will turn AI initiatives into real value, Reis adds.

The market has really bought into the GenAI hype, he says. “There is a belief that AI will do everything. But you can’t have a solution looking for a problem – you need a problem looking for a solution.

“So we are seeing companies demanding GenAI use cases where they don’t even have classic machine learning (ML) deployed yet.”

Yes, there are some use cases that can easily be addressed with GenAI. These include mostly instances where video, image or text content is used. But other use cases, such as tabular data or open ended models are still hard to solve with AI.

At the end of the day, he stresses, the value of any AI or ML solution comes down to the quality of the data.

“Data has got to be seen as the competitive advantage,” Reis says. “But just having more data is not a prerequisite. It’s the utilisation that is more important; the ability to combine domain expertise with data to solve problems.”

Harking back to the era of data lakes that quickly turned into data swamps, he says organisations need to remember the lessons learned there, and focus on data management.