Artificial intelligence is very effective at answering questions, based on the data upon which it was trained. It is, however, still quite limited in what it can do.

For instance, you could ask AI to book a flight reservation to Paris, with a seat in economy plus, and to give you an itinerary once you arrive. AI will list out all the steps to get those things done, but cannot yet of itself carry out those steps.

This is where cognitive computing comes in, with the ability to mimic human-like behavior. Cognitive computing is not a new term or practice, having been coined in the 2010s, but the advances in technology since then have moved it to the foreground. These platforms bring together such things as AI and ML as well as speech and object recognition and natural language processing.

“You can start to stitch the natural language understanding that these [AI] models bring to the table with the traditional programming that we are able to carry out workflows and actions based on those steps,” said Pankaj Chawla, chief innovation office at 3Pillar, a modern application strategy, design, and engineering firm. “That’s where the power comes into play. And you will start to see systems that are more and more human, in the sense that [those systems] can do more than one thing.”

Many organizations are already successfully using AI and have been for years, such as the Netflix or YouTube recommendation engines, or the Gmail spam filters. Chawla explained that those are more traditional AI – they’re fed data and are trained to recognize patterns – as opposed to generative AI, which is newer. “I worked at a company where they were employing humans to weed through bad posts and illegal posts and child pornography and things like that,” he said. “There was a lot of grunt work, not good work, that humans had to do, that you could relegate AI to do that work. Moving forward, I see the level of automation and generative nature just expanding and getting better and coming close to more of an artificial general intelligence type capabilities.”

Chawla cited advancements in technology from simple applications running on desktops to multiple always-on systems, and composable applications where you can, for instance, connect a billing system to a mapping system and a feedback system, and create business models such as Uber which never existed before. 

To the point that many fear AI might replace people down the road, Chawla said these systems will not replace humans. “All it does is it gives humans the ability to do a lot more a lot faster, at a higher order of magnitude.” Cognitive computing, he pointed out, represents the art of what is possible under the hood. “Blending the AI services, AI capabilities with the underlying data and product engineering capabilities that can bring about next-gen software systems is what I’m super excited about.”