DeepMind's CEO outlined the company's 2 major areas of innovation

Advertisement

DeepMind Demis Hassabis

Google DeepMind

DeepMind CEO Demis Hassabis.

DeepMind, a London-based research lab that was acquired by Google for £400 million, isn't just focusing on developing increasingly sophisticated forms of artificial intelligence.

Advertisement

The company's cofounder and CEO, Demis Hassabis, explained that DeepMind is also experimenting with the look and feel of its office, which sits within a huge new Google building in King's Cross.

"At DeepMind I always say we're doing two experiments," he said on Wednesday as he picked up Wired's Leadership in Innovation Award. "One is obviously working on AI and trying to push the boundaries of that. The other thing that we take just as seriously is trying to innovate on our culture and environment to try and create the best possible environment for scientific breakthroughs and innovation to happen.

Complimentary Tech Event
Transform talent with learning that works
Capability development is critical for businesses who want to push the envelope of innovation.Discover how business leaders are strategizing around building talent capabilities and empowering employee transformation.Know More

"We're doing all sorts of experiments on that all the way from the project management to the way we organise our building. So far it seems to be going pretty well. We've had three Nature papers in the last 18 months."

DeepMind, which now sits under Alphabet, Google's parent company, used to have its own discrete office in King's Cross before it moved into the larger Google office, which has enough room for up to 4,000 people.

Advertisement

The company has two dedicated floors in the 11-storey 6 Pancras Square, which have been given their own look and feel. Google is currently in the process of kitting out the building with massage rooms and a huge gym with a 90m running track.

6 Pancras Square

BNP Paribas

Google's new office in King's Cross.

DeepMind has made headlines over the last couple of years for creating self-learning algorithms that can master computer games and board games, anticipate energy requirements in data centres, and spot certain types of cancers.

Hassabis' organisation, which employs over 250 people, is split roughly down the middle into "research" and "applied", with Hassabis himself still contributing to many of the company's academic papers.