In the last couple of months I've been getting phone calls from fake charities using some sophisticated machinery. I assume it's all about economics from their point of view, if their hit rate isn't high enough to make it worth hiring English-speaking people, and they don't get good responses when the callers are poor wretches in India (though the "Your Windows computer has a problem" gurus there are still sharing their karma occasionally).

A couple of calls have come from "Mark Anderson for the National Patrolmen's Association". Before hanging up on the conversation, I asked "Are you a recording?" and he responded "No, I'm a real live human being". Later I thought, that machine actually has voice recognition, they're throwing technology at us.

So then a couple of days ago it was "Susan from the Women's Cancer Fund". She asked for my wife by name, and when I said "Sorry, she's out" the voice said "Oh, well I can talk to either of you", and I realized it was a scam. This time I thought faster and wondered what the computer would do if it got a question it couldn't have anticipated. Remembering the old joke:
I've invented a machine that can answer any question!
That's great. Ask it where I left my keys.
I asked "Where are my keys?" and the machine said "Oh, we're not really supposed to get into that." So then at the next pause I asked "What's the capital of the United States?" and this time I got "We could talk about this all day, but I'm calling about..." and at that point I hung up. They're probably doing the right thing, not wasting much time on off-topic stuff, but making some effort to keep up the pretense that you're being robbed by a real human. It's got beyond just launching a recording at us.

 

It's not AI. A real AI would have told you to check the sofa for the keys.

 

The question to ask them is why they're breaking the law to call you in the first place.

https://www.consumer.ftc.gov/articles/0108-national-do-not-call-registry

 

They've got methods to call without being traced, and methods to collect money without that being traced either. So if there's cash to be made and no enforcement, the law can say what it wants and they won't care.

 

They've got methods to call without being traced, and methods to collect money without that being traced either. So if there's cash to be made and no enforcement, the law can say what it wants and they won't care.

Yup. Anything beyond hanging up is just for fun.

 

So, if you have call forwarding, just have the robot talk to another robot. http://www.jollyrogertelco.com/

 

Answer "no" to ANYTHING asked.
It will be obvious you hear and understand any questions or statement.
They, or it, will soon tire of it.


Quick version, hang up. Block calls from that number.

I'm blocking over 100 numbers in my phone currently.

 

I got one last week. Dead line initially, I waited, said "hello" a second time and after 3 or 4 seconds I got a female voice that said, "oh, excuse me, I was just adjusting my headset. I'm sorry for the delay." Then she went on to tell me about the patrolmen's fund that I should donate to. The voice was a bit lacing in emotion or stops for breath so I suspected a robot. I thought the human touch of "adjusting my headset" was cleaver but could have been done more convincingly.

I got similar evasive responses from the robot like @John P did when I asked what number she was calling, what person did she want to talk to and more. It eventually just said "I will call back again at a more convenient time" and hung up in me. My fun ended.

 

The calls I get are annoyingly human - some guy with a thick Indian accent trying to sell me some life-alert thing I think it was. Maybe they tuned the AI to be barely intelligible, to sound extra real.

 

I've had the "adjusting my headset" one too. It's clever, adding a little touch that establishes the voice as human. Just a few weeks ago we were talking about the "Eliza" program and similar early pseudo-AI programs, where they'd put in a few canned responses to particular inputs that actually seemed quite real. Like you'd say "Mom liked my brother more than me" and the program said "Tell me more about your family". It doesn't take much, if you aren't being wary.

 

Next time use the Kirk Method to deal with the issue. Guaranteed to fry their computer.

 

I got an " adjusting my headset" call last week all excited about the weeklong vacation I was eligible for by staying at some unnamed hotel.

When I could not get a word in edgewise I just hung up.

That is probably a mistake. My favorite approach with my numerous Microsoft Technical center calls is to feign concern, put them on hold, put my phone down, and go about my business.

The more time they waste on me the less time to talk with someone who may believe them.

 

My favorite approach with my numerous Microsoft Technical center calls is to feign concern, put them on hold, put my phone down, and go about my business.

The more time they waste on me the less time to talk with someone who may believe them.

I feel exactly the same way. I once kept a phone abuser on the line so long that she said _she_ had to end the call because they were not allowed to call after 9:00 PM.

 

Just another data point. I also got the "adjusting headset" call. If I get more that a second of silence, I hang up. Used to play along with the #fakeMicrosoft guys.

 

Hi,

Are there any good call blocking boxes out there?
What about the numbers that change. I get calls from the same company but with different numbers sometimes.
Maybe we need to invent something for this. Microcontroller based.

 

Perhaps an inline processor that does the same thing the scammers do. Wait to hear a conversation and if one is not detected within a threshold (which should be dynamic. So as to adjust to the scammers adjustments, hang up. Otherwise pass the call through. It can send a reply, as the scammers algorithm still requires a delay before he respond. That way, a human will be told what you're doing and will be patient.

Once a phone has been blocked, web services could be used to generate a regional profile for use by law enforcement.

It could be enhanced further with voice recognition which would identify scam lines in use. Like, my headset needed adjusting.

I think this forum could collaborate on such a device. I am sure it would be a commercial hit. We could go on shark tank!

 

Perhaps an inline processor that does the same thing the scammers do. Wait to hear a conversation and if one is not detected within a threshold (which should be dynamic. So as to adjust to the scammers adjustments, hang up. Otherwise pass the call through. It can send a reply, as the scammers algorithm still requires a delay before he respond. That way, a human will be told what you're doing and will be patient.

Once a phone has been blocked, web services could be used to generate a regional profile for use by law enforcement.

It could be enhanced further with voice recognition which would identify scam lines in use. Like, my headset needed adjusting.

I think this forum could collaborate on such a device. I am sure it would be a commercial hit. We could go on shark tank!

Hi,

Yeah, would be fun

I guess there are a lot of things that could be incorporated into the device.
I once envisioned an automated system that could detect the false caller and give them a nice long menu to choose from with lots and lots of options, like a lot of companies have the nerve to do to their regular customers!
For a shorter example:
"Listen to the following as are options have changed".
"If you know the parties extension you can enter it now".
"For customer service, press 1"
"For returns, press 2".
"For..."
etc.

then when they press something send them to another menu.
If it is an automated system, they may not even be able to follow the menu or press a button.
Would be funny.
Could give friends and family a special code to enter that would push them through.