The next generation of computing devices may dispense with display screens and rely strictly on voice recognition, an engineer at U.S. search giant Google says.

Google has seen dramatic growth of voice search and recognition tools, with recent improvements encouraging more people to start using the capabilities, the company's senior engineering director Scott Huffman said.

"We are looking to a future where we have a whole variety of devices," Huffman told Britain's The Guardian from Google's London offices. "We have a super computer in our pockets, but also one in our watch, one in our glasses, maybe on our lapel as well as our laptop.

"Some of those have a screen and a keyboard but some won't, and we're seeing dramatic growth in the numbers of people interacting through voice recognition."

Google has already made its first foray into the world of voice recognition with its pilot project for Google Glass, its wearable computer offering navigation directions, photo and video shooting and basic web searches accessed mostly through voice commands.

The company is also expanding voice recognition services on its Android mobile operating system, running on almost 60 percent of the world's smartphones.

"It's a cultural thing, getting used to a technology like this, but there will be a whole set of devices built around it," Huffman said. "Two to three years ago we crossed the line from a demo to a real product, so now at least most of the time it can understand what I say. We're still work on the recognition of accents, and there are a litany of problems. But when it works, it is magic."