by Esan Swan
When two of the biggest internet companies in the world are fighting over you, you know you’re doing something right.
That was the case for artificial intelligence (AI) company DeepMind in 2014, after Facebook and Google’s parent company Alphabet both made offers to buy the start-up. When their deal fell apart with Facebook, Google snatched DeepMind up for the hefty price tag of a reported £400mln.
Since then, DeepMind has been a leader in AI technology and its application in the real world. The company has offices in Canada and the USA, but its main headquarters are in London: the financial capital of the world.
London has three times the number of corporate headquarters than anywhere else in Europe and DeepMind’s head office is only a few minutes from the main government offices in Whitehall. This gives them a direct advantage over their competitors both in the UK and abroad.
“Being in London brings a set of benefits that give us an advantage — which is really helpful when you’re working on something this important,” says co-founder and CEO Dr Demis Hassabis.
Hassabis, a former child chess prodigy, started the video game company Elixir after gaining a double first in computer science at the University of Cambridge. At the age of 17, he coded a multimillion-selling simulation game called Theme Park.
He finished a cognitive neuroscience degree at University College London and then went on to do postdoctoral degrees at MIT and Harvard. Other co-founders include the company’s Head of Applied AI Mustafa Suleyman and Chief Scientist Dr Shane Legg.
Since its inception, DeepMind’s mission has been to find intelligent solutions to the world’s problems, and in most circumstances it has.
In 2016, it announced WaveNet, a new neural network for generating raw audio waveforms that produced better and more realistic speech than previous techniques.
The following year, DeepMind pushed the boundaries of AI with its computer programme, AlphaGo. Now retired, the programme learnt the ancient Chinese board game Go faster than a human being and ended up beating the world’s highest ranked Go player Ke Jei in a three game match.
In May, DeepMind announced it had taught an AI bot how to navigate on its own using digital grid cells. The bot learned shortcuts and the quickest routes, once again beating out a human’s ability.
Past Controversy And Questionable Steps
But along with its mission to “solve intelligence”, the Google company has taken
some questionable steps and caused its share of controversy.
In July 2017, the UK’s Information Commission ruled that DeepMind’s test app had broken the law when an NHS hospital did not properly disclose how patient information was going to be used. The same report cites that the NHS hospital gave DeepMind 1.6 million patient data records as it was conducting research surrounding acute kidney injuries on an app called Stream.
Although not found responsible for any wrongdoing themselves, DeepMind reacted in a blog post saying that they should have been more specific about their project with the NHS.
“We made a mistake in not publicising our work when it first began in 2015, so we’ve proactively announced and published the contracts for our subsequent NHS partnership,” the blog post read.
Almost a year after this scandal, Prime Minister Theresa May pledged millions of pounds of government funds to artificial intelligence developments in the diagnoses of cancer and chronic diseases during a major speech on science and technology.
“Our goal is that three in four people will survive their cancer by 2034 and we support efforts that will help us achieve this ambition,” she says.
It’s not clear how AI and the technology underpinning algorithms and machine learning will affect the human experience years from now. What is clear is that DeepMind will be at the forefront, pushing to find more solutions.
Featured Image Provided by DeepMind