The US congress now has an Artificial Intelligence “Caucus” considering regulating how Artificial Intelligence (AI).  One the items they are looking into is figuring out if citizens should have a right to know that they are talking / chatting with an Artificially Intelligent piece of software or a human.

Many people feel tricked when they find out that they have been talking or chatting with an AI when the default assumption has been that people are talking or chatting with human representatives of the company in question.  Given the situation today and the obvious fact that AI conversations are going to become more and more human like, it is understandable that governments want to consider the implications.

This 8 minute video covers the FUTURE OF AI ACT that has just been introduced in the US Congress, which focuses on the military and is more broader than our narrow discussion about rights, but it does give you a sense of what is being considered.

https://www.youtube.com/watch?v=_hTQEqB5y9c

After some careful consideration however, the only scenario we could come up with in which a human really needs to understand that they are not talking to a human representative, was emergency services like 911.  The argument with emergency services is simply that during a crisis (shooting, heart attack…) there may be nuances (tone, slurred speech…) that a human can take from a conversation that AI’s cannot.

We are not suggesting that AI’s can not be very useful in emergency services communications (think about alerts, routing… that can all be done much faster, more accurately and more calmly by an AI than a human).  We are suggesting that in real emergencies human callers should know if they are talking to an AI or another human.

Other than emergency services, we can find little difference between an AI bot and a human as far are your rights go.  Both an AI and human will document your conversation and most will actually record the voice or chat conversation, so there are no privacy issues there.  An AI will always follow the privacy rules coded into it but a human may not and that could result in disclosure of your sensitive information, so in that regard and AI should actually provide more protection.

expert-system-flow-chart-exampleBefore we all jumped on the AI bandwagon, we used to have “Expert Systems” which followed flow chart based rules to route you to the correct answer you were looking for.  For example a large cable company might , press/say 1 for English, then press 3 for accounts, then press 1 for deactivation, then press 1 to confirm your address, then press 1 to confirm account deactivation.

In simple terms, where Artificial Intelligence differs from Expert Systems is in flexibility.  AI’s can ‘learn’ or at least adapt, whereas expert systems are ridged.

We expect the market to clear itself on this one as companies respond to their customers desire to know who/what they are talking to.  We expect to hear phone calls and chat windows to start with something like “Hi!  My name is Ann and I will be your personal assistant (code for AI computer software) that will help you get the right answer fast.”

The travel industry is already considering self regulation to deal with this issue so governments do not have to:

“As we operate in an industry that is incredibly personal, emotional and complex, maintaining the right balance between genuine human interaction and efficient automation is something we’re always trying to fine-tune and optimize throughout every stage of the consumer journey,” said James Waters, Booking.com’s global director of customer service.

On a personal note, I want to know if I am talking to an AI or not, but the question here is do you have a RIGHT to know that.   What do you think?

By John Roberts and Ian Matthews


0 Comments

Leave a Reply

Avatar placeholder

Your email address will not be published. Required fields are marked *