Language-based tasks seen as main corporate application for AI

A KEY application for artificial intelligence (AI) is automating language-based tasks, participants said at the Management Association of the Philippines 2nd ICT Summit.

“Anything from drafting reports, writing and responding to e-mails, but anything language-based, these advanced large-language models are extraordinarily good at helping you,” Scott McLiver, partner and Asia-Pacific leader in Generative AI, PwC New Zealand told a panel on Wednesday.

Suresh Eswaran, CTO of Hybrid Cloud Services, ASEAN at IBM Consulting, said that AI always should be viewed as a productivity tool rather than a means of replacing jobs.

“AI gives you a very good framework based on which you can implement your own reports that you want to do. It gives a huge productivity improvement, rather than starting from scratch,” Mr. Eswaran said.

He cited a Harvard Business School study that found a 40% performance improvement for a highly skilled worker compared with those who don’t use GenAI.

“AI can automate the creation of visualizations and integrate the data so that everyone in the organization can understand and see the same data,” Charles Patrick Ty, enterprise account executive of Tableau, said.

However, in times of AI hallucination, in which a large language model produces incorrect or nonsensical outputs due to insufficient training data, Mr. McLiver highlighted the need for a human in the loop to test.

“The big tech companies are working hard on these models only being creative when we want them to be creative and that’s the almost trillion-dollar problem that they’re trying to solve,” Mr. McLiver said.

Meanwhile, Mr. Eswaran added the integration and benefit of AI expands to cybersecurity, where it allows the user to start looking at anomaly patterns.

“It can keep up with any human watching for threats,” Mr. McLiver said, noting that AI has been used to reinforce security. — Aubrey Rose A. Inosante