Robot ethics

A couple of days ago, Sundar Pichai, the CEO of Google, got the kind of reaction that every hi-tech visionary craves when s/he demonstrates cool new technology.   But amid the clamor of oohs and aahs and whoops and hollers, we can just barely discern the uneasy murmur of ethics qualms.

For several years, Android smart phones have had Google Assistant, which is Google's version of Siri.  Ask it a question and it will (usually) answer.  I use it quite a bit for driving, not only for directions but to find the fastest route to familiar destinations; and beyond automobile navigation, it can find and present all sorts of information.  It's essentially the realization of the Star Trek "Computer".

Earlier this week, Pichai demo'd a new feature of Google Assistant, called Google Duplex.  Duplex can do more than respond to simple commands; it can hold actual conversations.  Pichai demonstrates two of them here - please click on the video and listen, if you're able to do so.

In the demos, Duplex makes a hair cut appointment, and then discusses the possibility of a restaurant reservation with a host for whom English is not her first language.  Even seemingly simple and brief conversations like those two can offer surprising complexity and subtlety, and Duplex seems to navigate them perfectly.

It's not only impressive, it's more than a little spooky how realistically the robot voice has been engineered.  There is no trace of "robotics" about it.  It inserts "er's" and "mm-hmms" at natural-sounding places, sprinkles in a little contemporary slang, and utilizes inflections that make it pretty much indistinguishable from the voice of a real person.

Is this capability an unalloyed good?  Matt O'Brien of the Associated Press raises some interesting questions:

Is it fair – or even legal – to trick people into talking to an AI system that effectively records all of its conversations? And while Google's demonstration highlighted the benign uses of conversational robots, what happens when spammers and scammers get hold of them?

Those strike me as a couple of pretty valid concerns.

Let me add this into the mix as well: O'Brien's questions are raised from the point of view of consumer rights, and that's a legitimate area of business ethics.  But let's also consider the impact on workers.  Whether it was inadvertent or clever, Pichai's two demonstrations were of Google Duplex making appointments on behalf of consumers.  But judging by the state of the technology, it seems likely enough that the roles in the two demo phone calls could be reversed: a human could be calling a robot receptionist or a robot restaurant host which also would manage the business's calendar and work assignments.  If it's cost-justifiable, that may allow these small businesses to fire their receptionists and hosts. 

That wouldn't be new news; technology and robots have been replacing human workers for years now.   But as jobs like supermarket cashier, service station attendant, bank teller and retail sales clerk have been made ever more obsolete, there aren't many possibilities left for people without a college degree or specialized training.  Customer service has been one of the few roles that continue to hire these workers.  I would say those days are numbered, too.  And taxi and Uber driver will be next as the self-driving autos continue to improve.

New tech is pretty cool, there is no doubt about it.  And lucrative: not only is it enriching young entrepreneurial types like Pichai, but it's transforming colleges; one commentator observed recently that the trend in higher education is that "universities are large STEM research centers with small liberal arts colleges attached to them."  But as O'Brien demonstrates, the dusty, fusty liberal arts have a role to play in contemporary society - at least until the robots start making ethical decisions for us, too.