DeepMind’s ethics board has been a closely guarded secret ever since the artificial intelligence company was acquired by Google in 2014. But a new report from Hal Hodson at The Economist sheds some light on how the ethics board came about and what it’s there for.
DeepMind and Google signed an agreement drawn up by lawyers in London called the “Ethics and Safety Review Agreement” in the year leading up to the acquisition, according to the report.
This agreement states that if DeepMind ever succeeds in its core mission of building artificial general intelligence (AGI)— sometimes described as a machine that can successfully complete any intellectual task that a human can, and widely thought of as the holy grail in AI — then the control of that machine will lie with those on a governing panel known as the Ethics Board, according to the report.
The Ethics Board essentially allows DeepMind to legally maintain a degree of control over the technology it creates, no matter how valuable or dangerous it becomes, according to the report.
One thing worth noting is a machine with AGI might not choose to listen to those that “control” it.
Despite repeated requests, DeepMind has never revealed who is on the Ethics Board, but The Economist reports that all three DeepMind founders (Demis Hassabis, Mustafa Suleyman, and Shane Legg) are all on it.
In December 2016, Suleyman said at a TechCrunch conference: “The ethics board is ongoing and it’s something that we have internally to oversee some of our efforts.”
He added: “We’ve always said that it’s going to be very much focused on full general-purpose learning systems and I think that’s very, very, far away. We’re decades and decades away from the kind of risks that the board initially envisaged. We’re putting in place a variety of other mechanisms that focus on the near term consequences.”
A year before that, at a conference held at Bloomberg’s London HQ in 2015, Suleyman said: “We will [publicise the names], but that isn’t the be-all and end-all. It’s one component of the whole apparatus.”
DeepMind is often described as being at the forefront of the AI race along with Facebook, Microsoft, OpenAI, Baidu, and several other firms.
So far, AI systems are only capable of beating humans at certain games and performing other niche tasks in specific fields, but they’re advancing quickly, with new breakthroughs being announced by major tech firms on a weekly basis.
DeepMind did not immediately respond to a request for comment.