What’s the point of AI if bots go bad?
This week at Olderhood we had the unusual experience of being contacted by a malfunctioning e-mail bot.
I say malfunctioning because, as best we can tell, the string of messages that we received were not dispatched by a human and certainly were not received by one. Although we promptly responded to the message at least three times, we continued to receive new ones inquiring as to why we had not responded to the company’s “amazing offer”.
The last message sent had a tone which was somewhere between cheeky and downright condescending. Worse, it then closed by adding that “in the absence of a reply to this e-mail, I’ll take it that I have your permission to close this mail trail”.
Surely this is just a silly mistake, I thought. I’ll just reach out to them using the chat feature on their website and find out what to do.
Apparently not. I was quickly informed that they only provided “tech support” on the web chat and they could not put me in touch with “Susan” (aka the bot).
We then tried reaching out to the company’s Facebook group moderator using the instant message feature, which was designated as “very responsive”.
I politely explained my dilemma again and, sure enough, I did receive an almost instant reply – but was informed that they had no way of putting me in touch with “Susan” and that I would “just have to wait for her to reply to my message”.
Not to be outdone, I then asked if they could put me in touch with someone else and, after some minutes of waiting, I was told that the job had most likely been given to someone else.
So, to recap … a company with the expressed ambition of being a global leader in AI (and an estimated valuation of more than $35 million if Google can be believed) has a rogue bot on its hands and two outsourced customer support divisions with no way of contacting someone inside the actual company.
And, to make matters worse, the customer service humans appear to be apologising for (perhaps even making excuses for) the company’s bad-girl bot.
No wonder the member of the leadership team that I reached out to on LinkedIn has not responded to my message (or perhaps this profile is just a bot too – I’m not really sure)!
In any case, the frustrating thing about this situation is that this all transpired with a company that we already do business with, and the project mentioned in the initial e-mail was one that we were actually interested in.
Is this an isolated incident?
Honestly, I’d like to think so, but I suspect otherwise.
The real message here is that it is irrelevant how powerful AI becomes, if it isn’t benefiting the humans it was created to serve it will be quickly dismissed as irrelevant.
Robin Trimingham is the managing director of The Olderhood Group Ltd and a business consultant, journalist, podcaster and thought leader in the fields of life transition and change management. Connect with Robin at https://bit.ly/3nSMlvc or email@example.com