DALLAS — AT&T is experimenting with artificial intelligence (AI) and combining it with software-defined networking (SDN) and network functions virtualization (NFV) to figure out a better way to run its network.

Speaking on a panel at the 5G North America conference, Brian Daly, director of core network and government regulatory for AT&T, said that the company is looking at AI as a way to operate its network more efficiently by using it to make decisions that currently might require human interaction today. “We see AI combined with SDN and NFV as a way to provide us with efficiencies that may not exist today,” Daly said.

But AI isn’t the only breakthrough technology carriers are looking at to make their networks more efficient. Monica Paolini, founder and president of consulting firm Senza Fili Consulting, said that machine learning is also an important tool that will help operators not only understand what their networks are doing today but also help predict how their networks will respond in the future.

For example, Paolini said that many Internet of Things (IoT) applications collect a lot of data. But the problem is no one has really figured out how to make that data effective. Machine learning is a possible solution. “You can collect data to predict the future in an effective way,” she said.

Artificial intelligence and machine learning are quickly becoming the hot new areas of focus for many companies. Google today announced that the company has formed a new group in the Google Cloud division to push into machine learning and AI.

Google’s AI efforts look similar to Microsoft’s push. Microsoft in September created the AI and Research Group, which consists of more than 5,000 computer scientists and engineers that will work on products and conduct research.

Plus, IoT heavyweight GE Digital today acquired a pair of machine learning startups as a way to add more capabilities to Predix, GE’s IoT platform. GE acquired Bit Stew Systems and Wise.io, both of which use machine learning to better process large sets of data.