Google’s intelligence team in the UK DeepMind made its first profit last year, losing £ 1 billion in 2019, the first sign that Google’s huge research revenue is starting to pay off.
The benefit of pre-tax returns comes after years a lot of losses since Google acquired DeepMind in 2014 for about $ 400m, and another $ 1.1bn loan that parent company Alphabet wrote in 2019.
DeepMind’s revenue, which comes from applying its expertise to the Alphabet business, has more than tripled in 2020 to $ 826m, according to his recent accounts at the Companies House. Staff and other salaries increased to £ 780m, up from £ 717m last year.
Spending a lot of money shows up the cost of talent while the big modern companies are rushing to develop AI technology. Stability university professors was received 10 times who paid for their training to work like Google, Facebook, Amazon and Microsoft.
DeepMind said: “During these reports we have made great strides in our research work to help find the scientific findings. Our explosions protein prediction was declared one of the most important achievements of AI in advancing science. ”
He also said that it was a matter of pride that their research was now “adding to the material and infrastructure” that is changing the lives of billions of people.
Google’s DeepMind team, which integrates its AI expertise into other aspects of the Alphabet, is divided between London and California and now has about 100 employees, mostly engineers, working “twelve jobs in a variety of fields including advertising, advertising. , shopping, YouTube, voice-speaking, cloud, infra and [self-driving car company] Waymo ”, according to LinkedIn staff history.
Although DeepMind’s revenue has risen steadily since it began to provide results, its sales are only available in Labels and are called “research and development services”.
Examples of the most recent commercial applications include a Google Maps partnership that helped fix “upcoming times” on Maps maps up to 50%, Google’s voice activator, and an Android battery saving project that now has more than 1bn active users, saving about 140bn minutes per month, according to the company.
The company is collaborating with foreign experts and moving to new locations where its AI expertise can show significant change. For example, last week published the results of a recent survey try with the UK Met Office, Britain’s domestic climate and climate work, using precise timing, location and intensity rainfall for up to two hours in the future.
DeepMind also reported its first release and energy use in 2020, citing total emissions of 643 tons of carbon dioxide. However, this only refers to the emissions from the use of electricity in offices and business trips for employees, rather than those created by the training environment in which machine learning machines are trained and operated.