Google’s AI demo at this year’s I/O sparks a whole new ethics debate

• Google Duplex is an AI experiment that allows the Assistant to make calls like a real person.
  
• Google is testing the tool right now.
  

Technology and ethics have always been at odds with each other. Ask around, many believe technology companies don't believe in ethics. Social media fuels that debate from time-to-time, with fake news, spreading extremism and more. But, an AI presentation by Google at its developer conference this year, may have sparked the ultimate debate.

On stage at Google I/O 2018, Sundar Pichai proudly showed his company's newest AI initiative. It's called Google Duplex and it allows the Google Assistant to make calls for you, booking tables, appointments and possibly more. Pichai stated that the technology is still under development, but the first instance makes it quite clear that Google is getting creepily good at artificial intelligence.

The ethical conundrum


When making the hair appointment for Lisa, the Assistant never tells the salon employee that she's speaking to a robot. Google told CNET that the Assistant will "likely tell the person on the other end of the line that he or she is talking to a digital personal assistant". This is of course when the feature actually rolls out to users.

That's an ethical dilemma right there. Experts have suggested possible misuse of this feature by marketers (to make unsolicited robocalls), political parties using this to make pitches and so on. Yet, that may not be the biggest problem.

After all, if a robot could convince you to vote for a party, it's likely that a human would too.

Hmm, umm, uh?

As countries scurry to regulate technology, Google's demo may give them food for thought. While the calls are obviously initiated by you, one could question who holds the responsibility for information shared over them. If your son sets an appointment using your phone, an appointment you aren’t aware of till the calendar notification rings, are you liable to honour it?

Moreover, when you make calls to local businesses, it’s data that can’t be easily obtained. However, when the Google Assistant makes the same call, it does so from a Google server. Who owns this data? Can Google access it willy-nilly? The company already has access to almost everything you do in a day, but should it also know each and every appointment you’re setting?

The Assistant won’t just make the call here, it will block time on your calendar. If done manually, a human could easily write, “hair appointment”, instead of mentioning the actual name and place of the salon. If Google can’t access your location at all times, it at least keeps some of your information private.

And last but not the least, what happens when businesses start using this method? Robotic sales calls aside, one wonders what would happen if the robot makes an error when it’s calling a customer. Would a business be liable for the robot’s error?