August 18: Censoring Updates
Hello!
Today, we will launch updates to our censor list. This will include a more robust list of phrases and words which should never be used in a profession, support giving setting.
Specifically, these will be geared towards:
- Reducing flirtatious comments & sexual advances
- Sharing personal information
Our goal is to continue to make 7 Cups a safe place for all users to interact! Thank you!
Okay, here's a new one. In reference to the apparently widespread site issues recently and many listeners' inability to respond to their listeners, I tried to ask Krink if "7Cups has the ability to email all the member/guests." The word "email" tripped the censor. I had to create an awkward rephrase, changing "email" to "write to," I think.
How annoying!
Trying to quash sharing of personal contact information and social media handles by purely technical means strikes me as a quixotic effort doomed to failure. I have seen this circumvented in chatrooms by methods as simple-minded as a series of posts of single digits or letters, which could never be filtered out in an automated way (you can't ban single characters or numbers!). Stopping that requires a quick-acting moderator.
Patently offensive words and phrases are good in a list, as they prevent casual offenses to people (that is, the sufficiently determined troll will surely defeat the censor every time; but someone who just isn't thinking will get that moment to reflect and hopefully be civil). But trying to get a censor to do much more only causes frustration and ridicule, at least absent some advance in artificial intelligence unlikely to be made by the site's developers!
Amen. I've gotten offensive messages from trolls who just change the spelling of the word in question by one letter.
I'm sorry it's annoying.
We are under no illusion that these filters can't be circumvented (locks on doors can be picked too, but they still serve a purpose).
What the filter does do is catch a lot of first-time negative behavior before the person has time to adjust to the filter, which can help establish a trail when attempting to defend against a willfully malicious person.
Please remember that many people are connecting through our service and not everyone's experience matches your own closely.
I would add that is an international chat and people can speak other languages.
I get censored from time to time, usually for the two reasons:
1) my English is bad, and my sentences can be bad formulated or I can use the bad combination of words, or expression.
2) I speak French, and some words can be bad words in English but be fully innocent in French. I was censored both in 1-on-1 and on my feed with innocent French words.
AAAAARRRGGGHH!
I now have two gang-rape survivors who are blaming themselves for what happened to them. I copied and pasted a quotation from a UK rape crisis website, and IT TRIPPED THE CENSOR. How am I supposed to have a full and meaningful conversation with these kids if I can't even copy and paste from the rape crisis websites?? Here is the quote I was trying to use:
Myth: Someone who has willingly drunk lots of alcohol or taken drugs shouldn't then complain about being raped.
Fact: In law, consent must be fully and freely given by someone with the capacity to do so. If a person is unconscious or incapacitated by alcohol or drugs, they are unable to give their consent to sex. Having sex with a person who is incapacitated through alcohol or drugs is therefore rape. No-one asks or deserves to be raped or sexually assaulted; 100% of the responsibility lies with the perpetrator.
That must be incredibly frustrating!
Yes, it is. The way things stand, I can refer these girls to the rape crisis center websites, but if I can't cut and paste from the sites I have no way to know that they are getting to the areas they need. Wouldn't it be better just to ban the trolls than to censor the volunteer listeners who are trying to help??
it would be better if it were that easy.
we can try to figure out what phrase is triggering the false-alarm there and work around it.
but banning trolls isn't as easy as it sounds, and the filter actually helps us catch a lot of trolls unawares.
Is there any way to determine whether most of the troll are guests or listeners? Because I'm finding my member/guests can say all sorts of offensive words without tripping a censor, and I'm curious to know whether trolls are willing to expend the time and effort necessary to become a listener.
it depends on the troll, but it's certainly much easier to troll as a guest or member.
with the number of services for creating free emails, and with the number of listeners accounts that some of them created, they know the answers to the 'training' by heart, and it takes just 5 more minutes for them to create an account. and they can be a lot more harmful, if it's what they are looking for...
so it really depends of kind of troll...
Has anyone considered how this can feel to a guest who is chatting for the first time?
I was talking to an anxious guest about a sensitive topic and they must have used a word which triggered the filter. They panicked, thinking they were going to be reported somehow and nearly disappeared on me.
How exactly is this improving the service we are giving to people needing help???
i. have. had. enough!!! of this censor. Just now it refused to allow me to post the following: "I hope that you no longer have the slightest inclination to think you were to blame for being raped." Krink, or Eric, or whoever the gods of 7 Cups may be - please get a grip on this! If I don't like how someone is talking to me, I can block the conversation. Struggling to help a young rape victim is hard enough as it is, but now I'm not even allowed to say the word "rape?" What gives???
@Kimmelane
sorry for the frustration.
@laura - can we look at what is tripping the censor in this phrase and reconsider?
@Kimmelane
Hi, there! :) Thanks for letting us know. My name is Jake! I am an ambassador here, if you didn't know. Thank you for being so kind to this member. It's true that you have done nothing wrong with this statement. I just tested this a bit and figured out what tripped the censor for you. The phrase "hope raped" is censored in the system. These two words, in that order, were in the phrase you said. We have this censor to prevent anyone from saying something like, "I hope you are raped." We would never want anyone to say something like that as it could be extremely hurtful, as I am sure you can imagine.
A way to reword that could be: "It's important that you know that you are at no blame for being raped." I know sometimes it is a little annoying to have to rephrase, but please know that it is to protect anyone seeking help here and to have the safest and warmest environment possible.
In addition, I have deleted the censored phrase from your account. Hope this helps a little! :) Thanks so much for all you do.
@Jake - I know you mean well and I thank you for that. However, rephrasing my sentence as you did creates one that is awkward, hard to digest, and powerless. Stopping a chat to figure out a clumsy rewrite destroys the flow both of thought and of conversation. This young woman is extremely traumatized - in fact, she attempted suicide two days ago - and the last thing she needs is a listener who can't even use the right words to describe what happened to her. Perhaps 7 Cups needs to focus more on ensuring that guests know they can block bad listeners and less on blocking their own good ones.
@Kimmelane
I understand the frustration but please realize that these are not either-or choices. We need to do all of these things to the best of our ability.
It strikes me as a fool's errand to try to anticipate all the word combinations that might form an offensive message, as well as to expect a listener, member or guest to sniff out ways to rephrase that do not run afoul of a censor utility whose inner workings are not transparent. Would this censor catch "h0pe raped" as well, which is identical in content yet crafted specifically to evade such censorship? Would it cover other deliberate misspellings?
I don't have a problem in principle with having some filters, but I don't think the technology is in place to profitably employ them for any purpose other than skimming away the low-hanging fruit of the most patently offensive individual words.
@squib
of course one can get around any censorship regime if careful enough and if that is one's goal.
still we will see the first time someone violates the rule and if they are trying repeatedly, even if they work out how to get around it after that, and so there are some benefits to such a filter.
@squib - I completely agree! All the censor is accomplishing is to drive off good listeners. Because, as you pointed out, anyone who wants to bypass the censor can do so by using the numeral 1 in place of the letter l, a zero in place of the letter o, and so forth. Simply omitting or scrambling letters will probably work too. Whyyyyyy make it hard for your volunteers while the harm you seek to avoid is not circumvented in the slightest? Because, trust me, a troll who is out to hurt someone is going to find a way to do it, but a good listener who keeps getting censored is eventually going to walk.
And the censor strikes again: I can't even complain to another listener effectively because I can't tell him the phrase that trips the censor.
@Kimmelane
I cc'd @laura on your original complaint so we should be able to adjust the censor to work around this problem.
@KrinkTheMellowUnicorn - thank you, Krink. I appreciate your efforts!
Well the mighty censors are back, got censored for saying "I hope you won't be bullied" hope to see that fix as this doesn't bring back good memories about this update.
@Ace13
Yeah, I ran into the same thing yesterday, saying something like "I hope you find a way to stop thinking you're worthless." It took me a long time to figure out why that was problematic! (it was one sentence within a fairly long message)
@Ace13 - hmm, i don't think anything changed recently but it does sound like there are a few more phrases to allow? cc @squib @laura
@KrinkTheMellowUnicorn @squib maybe it doesn't record what's in the middle and think we are saying "you're worthless" or " I hope you'll be bullied" or such thing
@Ace13
Yes, I'm certain that's what it does. After the fact I figured out which was the offending sentence and the fact that objected to a combination preceded by a censor-ignored "not."
It's just frustrating not to be able to speak in what for me is a natural way, in saying genuinely supportive things.
@squib I totally understand the frustration, I'd be lying if I told you I never got frustrated about this "update" hopefully it will be tweaked really soon
@Ace13 @squib maybe report directly to community team via support ticket to make sure it gets followe up on.