The robocallers have become sentient

I got a call from the conservative party of Canada a few minutes ago, on our land line. The number was in my local area code.

I’m translating into English below, but this took place in French. The guy was a mature man with a deep voice and spoke common Québec French, with the tone of someone reading a script.

Me: Hello?

Him: Hello?

(silence for 2 seconds)

Me: I’m listening.

Him: Oh, sorry My name is XXX, I am calling on behalf of the conservative party of Canada. I have a short survey, it will only be three questions. (short pause) Do you intend to vote for the conservative party of Canada in the upcoming election?

Me: No.

Him: In the context of this survey, do you intend to vote for the liberals or for the NDP?

Me: The liberals.

(Note: Actually, I’m undecided, but I’m not gonna tell the truth to that guy. Especially since this question doesn’t have an “undecided” option, and is ignoring three other parties that are regularly in the news.)

Him: Can you repeat, please?

(This is when I understood, with some stupor, that this perfectly human caller was actually a robot.)

Me: The liberals.

Him: As you know, federal taxes have increased under the liberals as compared to the previous conservative government. …

(Note: I’m not even sure that’s true.)

Him: … Do you think you are paying too much in federal taxes?

Me: No.

Him: Thank you for your time. (click)

If it hadn’t made me repeat I never would have known it wasn’t human. Extremely smooth, no clicks, very realistic tone and timing. There was that pause at the beginning, but this was probably meant to weed out those who don’t want to respond; it was handled very smoothly with the “Sorry” sentence that made it appear like an audio glitch.

This is scary stuff. This is the same party that sent out robocalls telling non-conservative voters that their polling station had been moved back in 2011.

The trick is to just not pick up the phone. Oh, and if you can swing it, get a phone number that is not local to your area. My cell number corresponds to Charleston, SC, with an 843 area code, but I haven’t lived there since 2007.

I see a number with anything like a Charleston area code (every now and then it will correspond to some other location in South Carolina) and I don’t even have to think about it.

Now then, can I tell you about our exciting advances in automated online posting? It’ll only take a moment of your time, and I’d be happy to carry on this conversation over the phone or via e-mail if you’ll just provide me with your full legal name, address, phone number, date of birth, and social security or other personally identifying government benefits number for non-US resident dopers.

Talk to you soon, and thanks!

Comment ID: 107794681&

Advances in speech recognition, replication and synthesis have been tremendous in the past couple of years; you should check out the Whats new, Atlas? thread if you’re curious. Advances in video synthesis are taking place at a rapid pace right now too. Couple all that with other advances in AI and by the 2024 election, I would wager, most people will be unable to tell if the person they are talking to on a cold call like yours are actually human or not.

Hercules, you were talking to a computer. A living, breathing, silicon-based lifeform. I call it a “dynamic robot.”

A computer with a (relatively) sophisticated parsing program, capable of determining your responses within a limited range. It can detect key phrases, and it has a default, fallback setting.

Detecting “Are you a computer?” triggers something like, “No, I just use one to help in my work!”, or detecting “fuck you” leads to a hangup.

This kind of computer is obviously cheap to operate, and in spite of the very, very low response rate, is very profitable. It fools many humans, and gathers many credit card numbers.

Sentient? I doubt if I’d call it that. More sophisticated than 10 years ago, yes. Just you wait for another 10 years.

Hm, that was a very telling mistake. I’m starting to suspect that Musicat is a robot programmed to translate Greek into Latin, and is just using its spare cycles to post here.

I wonder what would’ve happened if you had said, “Please.”

Just wanted to link to my recent link: a bank has “been testing and using Watson for a year, but is only going public about it on Thursday … a virtual banker it calls Reggie. It speaks directly with customers about simple matters, such as updating an address or ordering a new card … Twenty-two percent of the time, calls are completely handled (or “fully contained” in the industry parlance) by Reggie.”

Aaaah, see, a robot wouldn’t make that all-too-human mistake!

Hahaha! I can assure you, [sir], neither [Musicat] nor I are robots, we just find them very handy in our regular postings.

Now then, can I tell you about our exciting advances in automated online posting? It’ll only take a moment of your time, and I’d be happy to carry on this conversation over the phone or via e-mail if you’ll just provide me with your full legal name, address, phone number, date of birth, and social security or other personally identifying government benefits number for non-US resident dopers.

Talk to you soon, and thanks!

Comment ID: 107794681&

I’d like to see what would happen if someone responded with completely irrelevant gibberish:
In the context of this survey, do you intend to vote for the liberals or for the NDP?

La Fontaine can get knotted!
I’d just like to hear the sound of the server running the thing crash & burn.

Ask it to compute the last digit of pi.

I’ve tried that. If it can’t grok, it defaults to one of several phrases, like “I didn’t understand. Could you say that again?” or “I don’t understand. We will call back later,” or just hangs up with or without a salutation. If it parses the word “computer,” it responds with the standard, “No, I’m not a computer, I just use one in my work,” as ASL alluded to.

I don’t go that far. I just ask it basic math questions. (For example, what is two plus two?)

I discovered that “Once upon a time, there were three bears…” works just as well. I don’t believe the parser’s vocabulary includes that phrase.

I understand that the Conservative Party doesn’t believe the Bloc or Greens will be major opponents for them. (Or the People’s Party, but in this case I think they’d be right.)

There is another possibility – it could be a person in a foreign country using a computer to play back prerecorded responses. Why? Because people don’t respond well to telemarketing calls from people with foreign accents. Even a human playing back canned responses wouldn’t be able to answer your irrelevant questions without revealing their real voice, because they don’t have canned responses prepared for them, except for a generic “I assure you I’m not a robot” response

Although that was 2013. It very well could be that AI has improved enough in the past six years to eliminate the need for a human to listen to the call and press buttons to play back the proper response.

I get these frequently. Usually it is a police & firefighter association soliciting donations. Lately, they always start off with, “May I speak with Joseph?” I always say, “This is Joseph” just to see if it throws them off. I know it is a scam because they go to great lengths to explain this is a non charitable organization, and donations are not tax deductible. When they get to the “Can you give to help our first responders?”, I always say, “No! A policeman shot my Pa!” I never get tired of that.

These robots are used by generic boiler rooms, not necessarily police or firefighter associations. I get many from medical scams selling braces, soliciting donations for breast cancer research, or peddling fake “surveys”.

Just keep in mind that 100% of these “dynamic” robots are fraudulent, as they all go to considerable lengths to disguise their real names, locations, and purposes. It’s impossible to obtain an IRS form 990 to see where their money really goes. No legitimate charity refuses that.

Wow! I hadn’t thought of that. It’s quite possible.

This would explain a robo-call grouping I got a few years ago.

I could tell that I was getting a robo caller, even though it reacted well to different responses.

So I decided to have a little fun. When I got the call again…

I meowed in response to the questions. I got a chuckle the first time, and the second time. The third time, I got a guy with a fairly thick accent who was really angry. Too bad, so sad.