The tension between Google and AI brain trust DeepMind has always been fascinating. To outline the relationship: Founded in 2010, DeepMind is home to some of the world’s leading AI researchers who consistently publish insightful academic papers. nature Front cover. Meanwhile, Google bought Deep Mind in 2014 and suffered huge losses. Really You want to squeeze money out of your juicy brain.
So the recent story about the two companies Wall street journal It’s really interesting. Here Parmy Olson reports that Google has ended years of negotiations between the two companies and ultimately rejected DeepMind’s petition for more independence.
According to Olson, DeepMind told employees that the “end of last month” meeting was over. One suggestion from the founders of Deep Mind is that “according to those familiar with the scheme, the reason why the powerful artificial intelligence they are working on should not be controlled by a single business entity” is that the company has the same legal structure as a non-profit organization. However, Google didn’t join it, and said to Deep Mind that it makes no sense considering how much money the company has spent on Deep Mind.
This conflict is not surprising. Google executives have repeatedly said that the company’s future lies in AI, and numerous news stories suggest that the mothership is pressing DeepMind to commercialize its work. This has led DeepMind’s research into projects that use research on Android to improve battery life and reduce energy costs in data centers, but the financial benefits of these efforts are not clear. Meanwhile, losses for British companies continue to increase. In 2019’s most recent published data, it hit £470 million (approximately $660 million). If Google wants its money’s worth, it can’t give Deep Mind something like a non-profit status.
Along with financial pressure, another debate between the two companies seems to be the backbone of ethical oversight. One of the most popular factors in Google’s acquisition of Deep Mind was its commitment to establishing an ethics committee to ensure that its technology is always distributed fairly. However, the exact nature and scope of this board, including who is on the board, has always been unclear. 2019 report Economist The board even said it has ownership of the artificial intelligence created by DeepMind (a term referring to AI that meets or exceeds human capabilities in a variety of tasks).
The state of this board is WSJHowever, Olson noted that DeepMind’s future work will be overseen by a separate ethics committee that “mostly has Google senior management. Olsen Twitter This is the Advanced Technology Review Board (ATRC). This is called Google’s “Best Review Board”.
In the given statement The Verge, A spokeswoman for DeepMind said: “Of course we have discussed and explored the various structures within the Alphabet group to find the best way to support long-term research missions to address intelligence. We can’t be proud to carry out this amazing mission while continuing our operational autonomy and the full support of Alphabet.”
Update, May 21 at 11:04am EST: The story has been updated so that the Ethics Committee WSJThe report of is ATRC.
Update, May 24th at 4:55am EST: The story has been updated with DeepMind’s statement.