The road to hell

New Yorker cartoon by Robert Leighton

For years, I've followed the writings of journalist Ian Brown. It's getting a little tougher now, since I deleted my Twitter account a few months ago, but Brown is worth searching out, for his deep-dive articles that take ages to read.

Last summer, he interviewed Geoffrey Hinton for the Globe and Mail, and I carried the article around in a cloth bag, intending to re-read it.

And being too scared to re-read it.

I suspect that, unless you have a G&M subscription, the above link to the article will lead to a paywall.  Here are some "highlights", in no particular order - you can click on the parts with smaller print to make it more legible: 

In the ensuing months, I've watched uneasily as aspects of AI encroach upon my daily life.  A small irritation is the rather useless summary that now appears at the top of my Google search results.  QI, the British panel show, tells me that if you add f***ing into your query - as in "f***ing Tony Award winners" - the AI summary will not appear.  I haven't quite had the nerve to try that yet. I can swear like a sailor, but I don't like writing it down as evidence against me.

Then there was the weekend before last, which involved a series of slip-ups and poor decisions, leaving me in a cold sweat, and all arising from my attempt to do the prudent thing.  The road to hell is paved with good intentions, y'understand.

This is a cautionary tale.  I'm reasonably bright; this could happen to you.

It all started with my checking Demeter's email.  Demeter has given up on devices in general, but I go into the Shaw Communications web site regularly to check on her church updates, and to retrieve messages from people who know her, but not me.

I hadn't checked for a few days, and there were two emails, duplicates, saying something about having to update the Shaw Web mailbox, because the current version was being shut down -  2 days ago.

Blue button saying "Update".  I pressed it, and then thought:  Wait a damn minute.  Check at the website first.

Couldn't find anything about this at the Shaw web site, so decided to "Google" it, to see if there were any news about a scam.  There was a query about the matter, which seemed to be in some sort of forum, which seemed to be connected to Shaw.

All of a sudden, a conversation form popped up, asking what my problem was.  It appeared to be from a male with a little blurb under his name stating how well he was rated for answering questions.  I laid out my problem, i.e. determining if the emails my mother received were genuine, even copying-and-pasting the text of the messages.  I noticed that a female picture had supplanted the photo of the guy I thought I was asking.

Then I noticed that she was restating everything I said.  When I was preparing to be a hospice volunteer years ago, they trained us in a technique called "active listening".  The idea was divorce yourself from responding to people in distress with your own experiences, and to check if you understood by restating what you thought you had heard.  We were cautioned to rephrase carefully, to avoid sounding like a parrot.

No such caution was being applied here.  She texted pretty much like a parrot - if parrots could text. I looked below the photo icon of the smiling lady, and saw:  "ChatBot".  Apparently, my query was being prepared for the male champion question-fielder, and he would respond, after I registered, for a refundable fee of $2.

I was getting impatient, but two dollars seemed a pittance, and I registered.

Then my phone harped a text notification.  It was Bank of Montreal:  BMO Security Alert: (name of the "answer supplying company") $66.00 (date) (last four digits of my card) Was this yours? Please reply Y or N.

Sixty-six dollars? They said two dollars, I thought to myself.  

I broke into a cold sweat.  

I'd given them my information.  

I called the number on the back of my card, in order to jump the hoops of yet more automation, so the artificial female BMO voice could give me my balance, and the dates and amounts of my previous few purchases.

Nothing unrecognisable.  No $66 deduction.

In agony, I pondered whether to stop my card, or report it lost, or something.

I decided to wait.  

During the next few days, my newsfeeds seemed to feature stories of other Canadians bilked of thousands of dollars, and their banks refusing to help, saying they'd voluntarily done business with the scammers.

It's been ten days.  Still recognising my purchases, and not seeing a terrifying leap in what I owe.

It's a Sword of Damocles, but I guess it's a learning experience.  But no more lessons from chatbots, please. Back in your cage.