Good Robot, Bad Robot, You Robot, iRobot, meRobot,

By Peter Bernstein
31 Oct 2017

The “fact” that we stand to lose the artificial intelligence (AI) race to the Chinese, Russians, North Koreans, et al seems omnipresent in tech media buzz these days. It has jumped into national discourse because of its centrality to cyber war preparedness. While I appreciate our need in the U.S. to “win,” I am circumspect about AI hyperventilation and its touch tool, BOTs. 

I seem to remember social media evangelists touting its ability to bring us all together, end world hunger and social injustice, and cure cancer. Just ask Facebook, Google and Twitter execs, who now must tell Congress about Russian abuse of their respective services to interfere in Western country elections how things are working out for them. Interesting how simple turning kumbaya into NYET was so easily executed. 

AI to the rescue?

The use and value of AI should concern everyone in UCC and beyond. Yes, AI leadership will be a core part of our innovation leadership tool chest and have profound impacts. Proof points already exist for example that contact center workflow offload to bots clearly can make companies more efficient and arguably more effective. My concern is that glossed over in the “win at all costs, trust us all is good” myopia of the tech world is that AI is not all upside, even when used in theory for positive business outcomes.

Nuclear fission enabled great good. It also enabled creation of weapons to destroy humankind. Indeed, the latter came before the former. It is a slippery slope we appear to be heading down. This is why we should be looking not just at the bright side of AI, and mute some of the tech industries’ incessant cheerleading. I fret about Bad Robot. I don’t mean the production company that has brought us such TV shows as Alias, Lost and Westworld. 

Here are just a few questions I ask myself:

  • Why is it that when I am online and need to access a company’s website or a portal that I am asked, “Are you a robot?” and then am instructed as to how I can authenticate that I am not when authentication is just typing in some letters and numbers with no biometrics required? 
  • Why is it that the reverse is not true? 
  • Shouldn’t I have the ability to know if I am interacting with real or artificial intelligence? 
  • Shouldn’t I have the opportunity to decline permission to interact with AI-based capabilities even if properly identified, and be connected to a human?
  • Why is there no service that uses AI so that my bots can do battle with yours?

I do not know about you, but the risk of having my identity snatched is very real. It probably will be thanks to Equifax. Be aware if you are not that in a year, when bad people know most subscriptions to free protection will expire. There will be a spike in identity theft.  Why our policy makers seem incapable of dealing with this is a topic for another day. Where bad bots come in is their ability to help steal my identity. I am not paranoid. The infamous “they” really are after me, and you 24/7/365. 

What I have instructed everyone I know is that any unidentifiable and validated number should go to voice mail. If you are crazy enough to pick up the phone never say “YES” to anything. A bot with associated software-based functionality lies in wait to turn that YES into very bad experiences both immediate and longer-term. 

The good news at least on the calling front is that the Federal Communications Commission (FCC) while understanding there is no silver bullet, along with interested industry parties including the telecom companies, is working on this. In fact, I urge readers individually or through your trade organizations to become familiar with various Notice of Proposed Rule Making and Notice of Inquiry proceedings and industry alliance efforts. These include such matters as the creation of a Do Not Originate Database; fixing challenges with blacklists and white lists; Call Authentication Trust Anchor and STIR/SHAKEN. These combined with the continuing transition from POTS to IP and IMS-based telephony will have major impacts as they speak to the routing and authentication of all voice calls.   

An additional problematic issue with bots surrounds the question of accountability. As bots and other software take over more and more tasks associated with human decision-making, and more importantly the exercise of human discretion in decision-making, “Oooops!” should not become an acceptable answer. This is an excuse already rearing its ugly head. Like whack-a- mole, it needs swatting. Watch the old movie War Games, if you want to have nightmares. It shows what happens when seemly uncontrollable AI-like capability subverts an originally utilitarian purpose. To use an old phrase, “software is eating the world!”

Permission Please!       

Years ago I mistakenly predicted we were moving toward a permission-based world. It seemed at the time that what was most valuable in the future online-centric world was going to be our identities and online behavior. It looked as if an enterprising entity would arise that understood this value to sellers and actually pay us for giving the outside world access to us. Mea culpa, I totally misread peoples willingness to give up their valuable possessions in favor of free functionality and stuff. The folks in Silicon Valley correctly figured out they could get advertising dollars without having to pay to unlock the doors to our value. In short, enter precise profiling, Big Data and analytics, customized target marketing and an expansion of the planes of vulnerability for hackers with their spoofing and bots. 

Understand that I am not a Luddite. I have a deep appreciation for the positive value of AI. I would like AI to be good for me. That means doing more than making responses to various types of inquiries sometimes faster and enabling commercial entities to cut costs. AI really on my side would allow me to sleep better. We all could use the proper application of AI on moving us back to a world where permissions matter and our value is fully protected and monetization is a win/win for buyers and not just sellers. I want an iBot/meBot. It would make UCC a safer and more trustworthy place to spend time - that other most valuable resource we cannot create more of.