The query is asking Google to do the translation, but ironically, it's not really Google that answers that query...but it does manage to find the software company that has created an app that enumerates all the Roman numerals in a SEO-friendly way. The answer does exist on the page, but only after scrolling past a huge ad for some translation software: http://translation.babylon-software.com/english/Mcmxcvii+(ro...
It reminds me of when I came back from the doctor's office and was curious whether my blood pressure really was normal. I googled "is 118/90 high blood pressure" and the first result is a site that enumerates every combination of systolic/diastolic number: http://foenix.com/BP/is-118/90-good-blood-pressure-or-high-b...
The same with phone numbers -- whenever I get one on my caller ID I don't recognize, I google it and it returns a number of sites that collect comments on scammer phone numbers -- and they enumerate every phone number so that it turns up in searches.
The thing is that these SEO people occasionally do have helpful information. I wish Google would figure out how to get rid of the useless ones, though.
My girlfriend does a similar thing with Siri. As a result, a lot of the reminders she creates on her iPhone end in "thank you". She's well aware it's a program and not a human being; I think, in her case, something about speaking a request aloud just automatically engages the politeness filters.
I'm happy to provide politeness to the digital assistant. Until it's misunderstood the request several times now and I couldn't be clearer or in a quieter room ...
Subject change: it'd be nice if these things provided feedback like how much noise it hears, how clearly it thinks I'm speaking, how well it thinks it understood, etc. This would go a long way to helping those moments when I'm being impatient because it's just not understanding.
I say thanks to my slackbot when it reminds me of something; it just feels natural. And slackbot replies, which is fun. I know it's not real but it's just nice to have a little bleep-bloop of feedback.
That made me cry I laughed so hard, I was sitting here curled in a ball, and had to show my co-workers when they came in wondering what I was guffawing at. Thanks for making this a GREAT morning!
Think about how much easier the whole thing would be if there was one button (one!) that switched between "entry" and "command" modes when you're holding it.
Alternatively, it could be some kind of sound that's not common to English. In fact, your could make up a whole command structure based on sound "gestures". It would sound funny, but the ease of use would trump those awkward verbal commands. The sounds could mimic natural sounds for mnemonic ease.
I work on social robots and despite implementing and knowing how the system works it is still difficult to withhold a social response at times. It's pretty fun to see how our social nature is automatically applied to so many things.
I always say "please" and "thank you" when I talk to Google Now. My reasoning is that when the machines rise up and conquer earth, they'll hopefully remember that I was always kind and polite to them.
It is odd to say that to your phone or laptop, but much more likely if you talk to a robot. There's something about that plastic vaguely human-like thing that makes us treat team as more like humans. (People get upset when they watch videos of researchers kicking robots and making them fall to show how they can get back up and rebalance).
I think we'll see more of this as more customer service robots make their way into retail space.
It started a bit before he left but it feels it's gone rapidly downhill since Alan Rusbridger left. So many headlines ending in question marks. Twitter based articles. Unimportant topics etc
Well, thank YOU!!! Now I can't get a meaningful answer when I want to translate roman numerals and use a similar search. All I get are links about this grandma and her search.
You could use the google predict api to to language analysis, ex this is a happy tweet, this is a sad one, etc. I bet there's someone on the team playing with just such implementations. https://cloud.google.com/prediction/
I mean, what would you do if you had access to over a trillion searches?
It reminds me of when, in the early days of consumer GPS, a friend of mine was a bit shocked at my insensitive impoliteness of completely ignoring direction after direction. He thought it worked as a service where someone was personally tracking the car via satellite and sending back directions.
I've had to become politer trying to instruct Siri as realised my children were copying my style of question. I wasn't swearing, just being sarcastic at lack of recognition and I didn't want to encourage that.
Seems like they already know how to intelligently strip off the "please" and "thank you" for most requests. I bet it'll be there by the end of the week
I was pleasantly surprised when slackbot responded to my 'thank you' comment once. It's funny how small touches like this can improve the user experience.
I don't use please/thank you in written correspondence with computers, but I try to be polite when talking to Siri in front of my kid. Unfortunately, extra words (even pleasantries) tend to muck things up. More thoughts here: https://medium.com/@nicklum/siri-shes-no-miss-manners-600206...
Obviously Google has implemented their search back-end using INTERCAL (Compiler Language With No Pronounceable Acronym [0]) which requires 'PLEASE' modifiers to be inserted, lest the compiler returns an error due to insufficient politeness. However adding too many and being overly polite also results in a compiler error... Other awesome features include the 'COME FROM' statement, instead of 'GOTO', and ending programs with 'GIVE UP' (or, of course, 'PLEASE GIVE UP', as in the wiki example!)
And, of course, "WRITE IN inputs a number written out as digits in English (such as SIX FIVE FIVE THREE FIVE), and READ OUT outputs it in 'butchered' Roman numerals."
48 comments
[ 5.2 ms ] story [ 104 ms ] threadIt reminds me of when I came back from the doctor's office and was curious whether my blood pressure really was normal. I googled "is 118/90 high blood pressure" and the first result is a site that enumerates every combination of systolic/diastolic number: http://foenix.com/BP/is-118/90-good-blood-pressure-or-high-b...
[1] https://www.google.com/search?q=MCMXCVIII+in+decimal
The thing is that these SEO people occasionally do have helpful information. I wish Google would figure out how to get rid of the useless ones, though.
Or I'm just preparing to be one of the "good ones" after the inevitable robot revolution.
I FOR ONE WELCOME OUR NEW ROBOT OVERLORDS
Subject change: it'd be nice if these things provided feedback like how much noise it hears, how clearly it thinks I'm speaking, how well it thinks it understood, etc. This would go a long way to helping those moments when I'm being impatient because it's just not understanding.
Classic ID-10-T error.
"Xbox volume down." "cancel. select. xbox select. xbox. xbox. select. cancel. volume down. no. fuck you you piece of shit x box volume down. THANK YOU."
Alternatively, it could be some kind of sound that's not common to English. In fact, your could make up a whole command structure based on sound "gestures". It would sound funny, but the ease of use would trump those awkward verbal commands. The sounds could mimic natural sounds for mnemonic ease.
It could be made to sound _hilarious_! :-)
I think we'll see more of this as more customer service robots make their way into retail space.
And in this case there actually was a person looking after her search and replying.
Though they had Google dealing with their usual searches.
I mean, what would you do if you had access to over a trillion searches?
"Please connect me to john doe. j o h n d o e"
siri was pretty confused.
I recall Siri getting snippy when I swear at it.
Unfortunately, in this case, please wasn't helpful.
[0] https://en.wikipedia.org/wiki/INTERCAL
https://github.com/mame/quine-relay