That puts me in mind of the beginning of How the World Was Saved in The Cyberiad by Stanislaw Lem…
1 Like
I believe I’ve heard snippets of this, and I think I recognize the trope. I’ll go ahead and buy it on audible and listen to it during my insanely boring commute that might as well be on rails but I have to actually drive.
2 Likes
Oh, absolutely. I cut it a lot of slack, I just think it’s interesting how wildly the experience can vary. Even in calls from people with plain American accents I rarely get enough to understand more than who called, and even that’s iffy.
2 Likes
What we need is a way to get a machine to do what the human brain does. We identify the context of not only the language within which to analyze spoken sounds, but also the dialectal speech pattern. The trick, I suspect, is to enable context filters that aren’t too rigid. Speech algorithms probably do something like that, albeit probably in a more fluid and less categorical fashion.
The trouble, I suspect, is that they’re a lot more literal and a lot less imaginative. The human brain will always be looking for word associations within accents it has trouble understanding. It will try to model the ideas behind the words to cut through the ambiguity. A machine might be able to play the association game - my understanding is that’s how IBM’s Watson works - but it’s probably a lot harder to get a machine to construct an internal model of the speaker’s own mind from which to extract intent, since we don’t actually know how the human brain does this.
I remember when I read Feersum Endjinn by Iain M. Banks, I initially had difficulty getting into the swing of the phonetically written segments. But before long, my brain had realigned to reading a new “dialect” of English, and I started looking forward to those parts because reading that way was fun. But part of it was understanding the narrative. When I came across a word that I had trouble identifying, I would think about what words made sense until I got it. If I still couldn’t grok it, I’d usually have an epiphany later in the story as to what it was supposed to be. So not only was I making contextual guesses, I was doing so retroactively. And yes, a machine can conceivably make contextual guesses, but doing so with the efficiency and intuition of the human brain is a taller order.
2 Likes
I really hated those parts of Feersum Endjinn, but you might like Will Self’s The Book of Dave if that’s your thing.
2 Likes
Here are two I got a couple days ago:
What is the mint second but I left in your time of your life easier so please which I know some of them under the Google got to express, but I’ll informatica didn’t come into that I could set it up going to go see if they beat you at your face if they those set up through some of those good ass yes.
Does America capital if you’re not maria letitia, so please return if some of your mondo that I can circle Cocoa works best but if someone could down and come in this I’d appreciate it the data Coleman and associates any field trip safe if that’s all set up as knowing that those good ass yes.
To be fair, they both turned out to have been speaking Spanish. Interestingly, Google knew to directly transcribe “Informatica” in all that.
Also interesting to note: Google consistently transcribes “Gracias” as “Good ass yes”
6 Likes
Thanks! At first I wondered why anyone would write that way. It seemed needlessly obtuse. But like an optical illusion that you finally see, once I got into it, it changed the way reading felt because I was more apt to go with the flow, thinking less about the language and more about the story behind it. The language became perfectly transparent, more so that the other parts of the book, because I wasn’t admiring the beauty of the prose, I was just listening to what it was saying. It’s sort of like the different feeling between reading a poem and listening to a song.
1 Like
Well, yeah. I let the robot connect me and then I wait for the first question from the scammer as my cue to say something along the lines of, “yeah hey that’s awesome TAKE ME OFF YOUR LIST and don’t ever ever call me again.” …and then I hang up. It’s mildly cathartic with the added potential bonus that they may actually take me off of their list (yes I know they’ll never take me off the list).
Or, if I’m feeling froggy, I will entertain myself at the expense of the caller. This takes several forms (and voices, sometimes) and is nothing more than my juvenile self reasserting it’s hold on my consciousness.
BAH BAH BOOEY MOTHERFUCKERRRRRR and suchlike.
3 Likes
The Cognitive Dissonance Podcast did a recurring segment using google voice transcripts of their voicemail inboxes for a while. The podcast ended up training google speech recognition to recognize the phrase “Gloryhole, motherfuckers” because that’s the fanclub’s standard salutation. ICP has (had?) one too “WhoopWHOOP!”
3 Likes
I got a call from one just the other day that appeared, according to the call information given on my Nexus, that I was calling myself. Which is dumb because I can just email myself and skip the telephone call 
But this is the part I don’t understand. Why is it still possible for these companies to begin the call with fake information? Why hasn’t this shortcut/hack been mitigated yet? And is there some way to mitigate that hack that I should know about (other than not answering the call)?
3 Likes
Here’s a classic…
5 Likes
What always struck me as peculiar, though, is the role of the telcos.
Caller ID is just a consumer toy, it was never intended to be difficult to spoof, and it sure isn’t. ANI, which is used for billing, is a different story.
Either these scammers are successfully stealing fairly massive amounts of service(and driving customers away) from the telcos, in which case you would expect some politically influential screaming to be happening; or they are being tolerated(presumably ‘pink contract’ style); in which case we should have a hard, but nicely concentrated, target to pursue in the fight against spammers.
3 Likes
I actually kept the transcript of one spam call for a long time just for shits & giggles because it was the most hilarious one I’ve ever gotten. Like @LDoBe, most of the time I am quite impressed with how well Google Voice does in the transcription, but when they get it wrong, they get it WRONG.
2 Likes
It’s weird. I was happy with the nadsat in Clockwork Orange, and the mockni in The Book of Dave (nobody reads Will Self for an easy read), but it grated in Endjinn. Just got in my way of reading, I felt.
(sweeping statement alert!) I don’t much like poetry.
1 Like
Guesses in no particular order:
-
Until a few years ago, most telcos in the US were regional monopolies. Then mobile phones gave consumers choices, and they started jumping ship. Perhaps they’re so used to not having to solve the problem that they can’t change their stripes now.
-
Maybe scammers do get disconnected, and then proceed to switch numbers. The wack-a-mole theory.
-
Legal immunity on the part of the telcos combined with too many offenders for them to pursue civil actions against, particularly since it’s hard for court servers to locate someone who obtains service under a false identity.
1 Like
Ah, I love it. To me the best prose approaches the lyrical*. But not necessarily the best storytelling. When writing is beautiful but not in direct service to enhancing the storytelling, it’s sometimes easy to focus on it and not what it’s about. Great poetry crafts words beautifully to tell a story. Jokes have the same appeal. A joke is basically a mini-story and a well written joke uses the words to enhance the story.
Here is, IMHO, a wonderful example of beautifully crafted wordplay that only serves to enhance what the poem is about:
*I also enjoy non-lyrical poetry, but differently and less frequently.
3 Likes
I totes want Phone Spiders!
3 Likes
No… but often I’ll wait for the opportunity to flush the commode, yell an obscenity, cut a flatus etc.
On a related note (well, I want to link to my website): here’s a call I received a few years ago.
5 Likes
Mr. Mabe is now my hero, and I’m sad I’ve never taken that approach before.
EDIT, GODAMNIT: Just got a call from (305) 258-7465, some sort of telemarketer based in South Florida (at least that’s their–and my old–area code) called me with a robomessage. Pretty sure it was “Card Services”. They ask for callees to press a number and be connected to their representative. I hit many numbers and was awarded with a robo-countdown of my progress in line, which went quite rapidly from #72 down to #1…and I blew my chance. Representative Fuq-Whad began his talk about saving me money and I jumped in with, “MAN I CAN’T FUCKEN WAIT to get my credit score tripled–gonna load up on benzos and puppies with that new credit card!”
And he hung up. I don’t feel like these telemarketers are really, truly, listening to me.
3 Likes
Surely proof that heroes aren’t born, they just answer the call!
Sorry, hard to resist that one.
3 Likes
