By Nicola Truth
Is a recent article posted on Futurism.com simply an interesting business story or is it a cautionary tale about what could be ahead for job boards?
Dan Robitzski’s piece recounted an incident related to DeepMind, an artificial intelligence research unit at Google. The company was acquired by Google in 2014 and is generally considered a leader in the global quest for artificial general intelligence or AGI. That’s the term used to designate true machine cognition, the capacity to acquire external signals and, without guidance, to analyze those signals and react appropriately (or at least as humans would).
Today’s state-of-the-art in AI is something far less than AGI. Indeed, it’s best described as the first of hundreds of generations of the technology that will be required to give machines the ability to think as humans do. That said, DeepMind is clearly making progress. Beyond teaching machines (such as its AlphaGo) to beat humans at complex games, its goal is to create “the best techniques from machine learning and systems neuroscience to build powerful general-purpose learning algorithms.”
What will robots and humanoids imbued with such general-purpose neural networks be able to do? A better question would be, what won’t they be able to do? They’ll be smarter than humans because they’ll be better at acquiring and analyzing data; they’ll have more common sense than humans because they’ll never forget a lesson learned; and they’ll be wiser than humans because they won’t (in theory at least) be swayed by prejudice or a preexisting opinion.
According to Robitzski’s article, when Google bought DeepMind, one of its founders, Demis Hassabis, established safeguards to ensure its independence from both Google and its parent, Alphabet, Inc. These protections were designed to prevent Google from latching onto DeepMind’s AI developments and using them however it saw fit. That’s clearly within the purview of an owner, so the concern was apparently less about what Google might do and more about how it would do it.
It appears Hassabis and his colleagues were worried that new products and services derived from their research would be launched without first considering the ethical considerations of doing so. As Hassabis put it in a 2015 interview with BBC, “I think artificial intelligence is like any powerful new technology… It has to be used responsibly. If it’s used irresponsibly it could do harm.” Based on that concern, they installed a mechanism – an ethics review board – to provide at least a measure of oversight and a way of considering the potential consequences, before unleashing anything into the public domain.
Two years later, DeepMind launched a new line of business for its AI products. It focused on serving the exploding healthcare market. By 2018, this new division was signing contracts with hospitals to improve their operations and patient care. Beating humans at GO was a feather in their cap; beating other companies in the market was a plus to their bottom line. The company seemed poised for significant growth, and no less important, it would achieve that success while its work was overseen by the ethics review board.
Then, something unexpected occurred. Within months of the announcement of DeepMind’s new contracts, Google launched its own entry into the healthcare market. Following a pattern it had used in other markets, the company gave the new unit a simple, all-encompassing name. It was called Google Health and targeted the exact same customers as those of its subsidiary, DeepMind. Not long after that, Robitzski reports, all pretense of a symbiotic relationship was dropped and Google Health “absorbed DeepMind’s program into it, giving little notice to DeepMind or the hospitals.”
What does this tale have to do with job boards?
This collaboration-into-competition trope should have a familiar ring to many sites. In the first decade of this century, they trusted their future to a partner that promised a shared pathway to success, only to have that partner later renege on the deal. While DeepMind’s situation is clearly different – it is a Google subsidiary, after all – the abrupt and total annexation of DeepMind’s growing healthcare business should raise a question job boards ignore at their own peril: will collaboration once again turn into competition in the jobs market?
The answer to that question is unknowable, at least for the moment. Therefore, there’s a second question that job boards should be asking: how do we protect ourselves just in the case the worst occurs?
The opinions expressed in this article are those of the author and do not reflect the views of TAtech: The Association for Talent Acquisition Solutions.