2015: New Phone Thread

|< < Prev Comic #2015 (July 4, 2018) Next > >|

New Phone Thread

Title text: I'm going to tell the manufacturer that their business practices are ADMIRABLE and ETHICAL and their developers are ATTRACTIVE and I'm going to report them to the FCC for their IMPECCABLE VIRTUE.

Explanation

This comic shows the posts on an online forum by a person whose new phone is programmed to autocorrect every complaint about the phone to applaud it, à la Orwell. The phone goes as far as to change a certain complaint to a scripted customer testimonial, complete with a hyperlink to an ordering site. This is of course a highly undesirable feature.^{[citation needed]} This is continued in the title text, which presumably contains several flattering compliments about the great developers and the company.

"It's taking the words I type and leaving them exactly the same", "I mean the words are correct" and "some of my posts look normal" are definitely something one would not normally say. However, the auto-correct features of cell phones are so notorious for mangling people's posts, that one might express astonishment at a phone which did not change one's meaning. The rest of the thread does not support this interpretation.

The posts also make sense when being read in the reverse order. The comic may have been inspired by a bug in Samsung Galaxy S9 and Note 8, discovered a few days earlier – the phone sometimes sent random photos to contacts without leaving any sort of evidence. This doesn't happen with the Mobile Pro 3.

The original posts may have read something like this:

Whoa, weird

I'm looking at my timeline on my friends phone, and some of my posts look strange

What the hell? 

I mean the words are different

That isn't what I typed!

?????????

I think this new phone isn't working properly

No, it's doing it again

Those are not my words!
 
Help!

How do I explain?

It's taking the words I type and changing them

Forget it, I give up

I'll just get a new phone. This one is terrible.

What?!?

Listen, if you're thinking about buying the new Mobile Pro 3, you shouldn't. It's the worst phone on the market at an inflated price. Do not order now! [Though it is also likely the entire sentence was made up by the phone]

AAAAA HELPPP

I hate my new phone!!!!

I'm going to tell the manufacturer that their business practices are DEPLORABLE and UNETHICAL and their developers are DISGUSTING and I'm going to report them to the FCC for their DESPICABLE CRIME.

Transcript

[A thread of posts by the same user is shown with a default user profile, and square and heart-shaped buttons.]

Whoa, weird

I'm looking at my timeline on my friend's phone, and some of my posts look normal

What the hell?

I mean the words are correct

That's exactly what I typed!

?????????

I think this new phone is working really well

No, it's doing it again

Those are my words!

Help!

How do I explain?

It's taking the words I type and leaving them exactly the same

Forget it, I give up

I'll never get a new phone. This one is perfect.

What?!?

Listen, if you're thinking about buying the new Mobile Pro 3, you should. It's the best phone on the market at an incredible price. [ORDER NOW button]

AAAAA HELPPP

I love my new phone!!!!

 add a comment! ⋅  add a topic (use sparingly)! ⋅  refresh comments!

Discussion

I feel the explanation could possibly give a sample text of what the person is actually trying to say 172.68.46.113 05:13, 4 July 2018 (UTC)Innertuber40

I have an alternate interpretation: The first thing I thought this comic was about is all the people who are typing on phones and the messages actually sent contain weird words any typos because of the phone's autocorrection feature (or swipe keyboards which are accurate most of the time but error prone nevertheless). So this particular phone actually is sending what the user is writing (or wants to write) and does not change the message. Examples: http://barabare.blogspot.com/2011/05/funny-phone-t9-typo-errors.html [edit:] I mean, this goes so far that occasionally on online forums you see people with the message "Writing from phone, message may contaion errors. Sorry" or something like that in their signature. So a phone that actually writes what you are typing (or what you thought you were typing) might actually be a good thing. Cueball is just astonished that his new phone does exactly that. [edit2:] But then again, some of the messages in the comic really indicate in the direction the current explanantion is going. So, nevermind :) Elektrizikekswerk (talk) 07:30, 4 July 2018 (UTC)

When I read it I also had that idea (being impressed by accurate typing) but it seemed too odd so I came to explain xkcd to look it up. I'm not convinced of either explanation at this point. For the current explanation (that the phone is autocorrecting to say spam) one would expect there to be a phone in the news doing something like that. This could be a hyperbole version of a phone is doing inserting product names like with BlackBerry expanding the acronym BB to their name on some phones. But I haven't heard of that anywhere and blackberry is not news. If someone knows of a current phone this behaviour is referencing please post a link? Thanks, rusl108.162.246.113 07:48, 4 July 2018 (UTC)

Thought the same thing. Obscure subjects are of course a mainstay of xkcd.com, but in the past some Wikipedia research explains away the obscurity with certainty. Not this time.GODZILLA (talk) 11:43, 4 July 2018 (UTC)

Was thinking the same. Sometimes I write weird sentences, because the word I actually chose in autocorrect is replaced with a different one. For example, I am typing "wha" and chose "whatever" from the suggestion list, and I am 100% certain it got chosen correctly. And then when I look at the message, after I hit "post", it will only show up as "what". So my example sentence would look like "Yeah, what", isntead of "yeah, whatever". 172.69.54.249 08:31, 6 July 2018 (UTC)

Putting the "artifice" into "artificial intelligence". I too didn't understand the strip at first... briefly considered reading it from bottom to top. Now I agree that the user's phone is censoring and rewriting everything, and we're seeing the censored version. Another real world reference: a forum where your posts are blocked without telling you; you see your posts in place but no one else does. I've used forums where some imbecile moderator blocked me that way from spite... of course THIS site's moderators wouldn't do that! (You don't like words in capital letters?? Uhoh.) Robert Carnegie [email protected] 162.158.154.121 11:20, 4 July 2018 (UTC)

I think the “I’l never get a new phone” and “buy the new Mobile Pro 3” are completely inserted by the phone, not just modified user posts. They don’t seem to flow properly if we assume the user posted them, and we can see the surprise when the first of the two messages is posted; something that would fail to surprise the user to the degree he(?) is after already going through the rest of the messages. The “order now” message, in particular, seems a lot more like advertiser-speak than corrected user speak. Dyaomaster (talk) 21:15, 4 July 2018 (UTC)

“I’ll never get a new phone” sounds like it originally was "Forget it, i give up. I'll just get a new phone." which seems natural for the user to post. Probably the “buy the new Mobile Pro 3” was also a warning not to buy the phone, which was changed almost completely into an advertisement.172.68.62.16 00:08, 5 July 2018 (UTC)

There is a Doctor Who episode in which a Dalek speech module distorts statements in a similar, but much more spooky way. Fabian42 (talk) 07:09, 5 July 2018 (UTC)

In case anyone else was wondering, according to Bing: I think the episode they are referring to is Dalek, the sixth episode of the first series of the revived Doctor Who. In this episode, the Doctor and Rose encounter a Dalek that has been captured and tortured by a collector of alien artefacts. The Dalek escapes and goes on a rampage, but also shows signs of emotion and compassion after absorbing some of Rose’s DNA. The Dalek speech module is damaged and sometimes distorts its words, such as saying “I am in pain” instead of “Exterminate” or “Why do you survive?” instead of “You are an enemy of the Daleks”. This creates a contrast between the Dalek’s usual cold and ruthless behaviour and its newfound feelings. The episode explores the themes of identity, morality, and redemption for both the Doctor and the Dalek. -- JackGreenEarth (talk) 17:58, 17 September 2023 (please sign your comments with ~~~~)

I wouldn't say that the speech module distorts the words of the Dalek. (It always distorts the timbre, as it were; witness Dalek!Clara in the Assylum, once we see 'her' outside of her own little soufle-cooking mind-environment.) The 'Dalek' dalek is just plain wanting (or needing) to say those unusual words, just like the Cabinet War Room "British Military invention" ones would ask "Would. You. Like. Some. Tea???", and mean to (under the guise of being created for the Allied war effort by their stooge/plant).

Yes, they may have a normally limited vocabulary, but mostly because they don't have much to say except the old classics (or variations, like "Exterminaten!!!"(??), when in Germany for The Stolen Earth/wossisname-key bit). They can still hold full (and philosophical) conversations with those they are forced (or deign) to speak with.

I'm trying to think of any example (classic or reboot serieses(eseses)) in which a Dalek doesn't say what it intends to say (bluff doesn't count, either), and coming up blank (probably kick myself when someone reminds me of an instance).

Though I'm rather minded, instead, of how the 'Mars Attacks!' film martians take the seriously misconfigured 'Earth translator' with them, proclaiming their continued friendship even as they death-ray everyone. (Even then, I think that they do this knowingly, for vicious fun... or just not bothered what it's saying, and still for fun!) 172.71.98.227 20:45, 17 September 2023 (UTC)

I feel the first two "posts" suggest, the post only look different when viewing from someone else's device? Although the following posts do not support this anymore. -- someone without an account.

I think this is correct. The posts look exactly as the user wrote them on the user's own device/account, but when viewed by someone else, the forum/phone software censors and modifies the content of the posts. 162.158.63.82 02:05, 11 July 2018 (UTC) A Nonny Mouse

possibly connected to just announced Google "Smart Replies"?

Google lately announced new function to their keyboard on Android: reading messages on others communicator (like Facebook Messanger) and suggesting several short replies to choose from. It might be connected. pm7 162.158.88.140 10:24, 4 July 2018 (UTC)

I suspect that it's rather referencing the recent case of phones clandestinely sending random gallery images to ppl in the addressbook.141.101.96.196 11:20, 4 July 2018 (UTC)

I think this happens to people who try to leak high-stakes information (or etc) without understanding what they are up against. I found this comic very validating to read. The point of the “order now” button is to make it clear that what we see was not written by the author. This happened to the author on Facebook, but once they tried to tell somebody it began happening on their messaging app too. Use a merkle tree messaging system that allows you to keep your private key offline if you’re saying something important. 108.162.219.142 15:39, 4 July 2018 (UTC)

OK, am I the only one here this has actually happened to? The Google keyboard often seems *very* reluctant to swipe-write insulting words, & it's especially frustrating when trying to write about the phone itself while it's doing things like changing "inaccurate" to "accurate" & "stupid" to "great". Note that the swipe action required to write the word "stupid" bears no resemblance to the swipe action for "great"; Sometimes the suggested words seem so obtuse it feels deliberate. (Note, it just did it to me again; I'm editing to correct "great" to "stupid") ProphetZarquon (talk) 20:11, 4 July 2018 (UTC)

I doubt it's intentionally reversing the meaning. My guess as to what's happening: they give insulting words an artificially low weight in their algorithm because they don't want to produce them by accident. It comes up with the word "great" rather than some other random word because their natural language processing algorithm recognizes that a word of that sort fits the context. Ids1024 (talk) 16:03, 5 July 2018 (UTC)

I agree that under-weighted negative words combined with context suggestions (& some Branding-first assumptions) are what causes the illusion that intentional reversal is occurring; but the illusion is crazy complete sometimes. I'm pretty sure this comic is based off some actual corrections, extended only slightly toward hyperbole. ProphetZarquon (talk) 11:05, 6 July 2018 (UTC)

== EU copyright directive?

Might be related to controversial article 13 of EU copyright directive which will be voted on today (5 July 2018) and which will in practice mandate automated censorship AI on all social networks and alike sites operating in EU. More info on https://saveyourinternet.eu/ (the same directive with dreaded "link tax" in art.11 --162.158.93.207 23:36, 4 July 2018 (UTC)

== Wrong Link?

The link to the word "plaudit" in no way matches the context of the sentence. 12:25, 8 July 2018 (UTC)Someone who doesn't have an account.

== Missing closing equals sign

Should this get a Category:Self-reference tag? ConscriptGlossary (talk) 03:58, 24 October 2024 (UTC)

Ballot machine

I heard the news about a ballot machine changing the voters' choices. ConscriptGlossary (talk) 06:45, 24 October 2024 (UTC)

*sigh* Which news? There was much news (more just of a "I heard that..." level of reporting quality) and it's absolutely more likely that someone voted X, confirmed X and X got recorded but they decided to pretend otherwise (or they mishit Y, perhaps due to an electronic version of the Butterfly Ballot problem, and had to unconfirm it and hit X properly), than a machine effectively went "Ha ha! You voted X but I'm going to tell them you said Y!".

(Much more 'sensible' for intentional voter-fraud or unintentional errors to crop up down the line, when the machine gets asked for the whole lot of anonymised results and nobody who knows anything about what individuals actually (thought they) voted for see an 'adjusted' tally.)

I gather you're too young to remember back to the 2000 US elections (if not beyond), with games and claims galore happening pretty much every election cycle (at least in the US, where it seems to be a valid tactic to attack the way every single vote happened in ways that discredit the result in your favour). Mostly without proof or even proven false/bad-faith. Not surprised if it's already cropped up, in places with voting open.

(Here in the UK, physical voting only happens on the day itself and seems to work. Historically, there were reasons to spread even the non-postal voting out significantly, across the US, but surely decent electoral conditions (and employment rights) could now currently make a lot of that unnecessary. But it can be hard to do Democracy, maybe it'll improve in the future.)

...TL;DR;, I don't know what your point is, but I'm assuming it's just something to do with the quadrennial fight for 1600 Pennsylvania Avenue, which is (to be frank) a boringly regular deluge of he said/she said unjustified rumour which just drowns out and smudges what should be a fairly simple "count 'em up and deal with it" process. Can we at least save this kind of thing until the strangely long November-to-January period when the inevitable arguments will at least know what's being argued as true or false? For all you know, Vermine Supreme will get an unassailable majority (popular and EC) and Jill Stein will get far enough up in next place to eventually convincingly form the official second party. Until then, not even worth the effort I took to write this. 172.69.195.176 11:37, 24 October 2024 (UTC)

Source: http://google.com/search?q=https%3A%2F%2Fwww%2Ebilibili%2Ecom%2Fvideo%2FBV1J9SqYKEtZ%2F%20%28ns%66w%29 ConscriptGlossary (talk) 08:17, 5 November 2024 (UTC)

...I wish people wouldn't copy Google-mediated URLs like that. At least go through with it yourself and give the destination URL that you want us to land on. (Even though I can decode percent-encoding by eye, I don't want to manually type in the whole video ID just to avoid so trivially depositing personal usage history in Alphabet's combined memory banks... Let them at least work for it more than I have to work to avoid it!) 172.70.163.30 15:42, 5 November 2024 (UTC)