Fend Off Trolls, Bots and Jerks With ‘Empathy’ Test

http://www.wired.com/threatlevel/2012/10/empathy-captcha/

by 001sky • 14 years ago

45 35 14 years ago

wired.com

starpilot• 14 years ago

Hmm. No involuntary dilation of the irises...

meepmorp• 14 years ago

I admit that I was disappointed there were no tortoises involved.

ktizo• 14 years ago

The tortoise lays on its back, its belly baking in the hot sun, beating its legs trying to turn itself over, but it can't. Not without your help. But you're not helping...

tnuc• 14 years ago

What's a tortoise?

coderdude• 14 years ago

Trolling is the art of feigning one or more of the following: empathy, ignorance, sincerity. You're really only fending off non-native speakers and spammers who give up very quickly (one could guess and ultimately get many correct answers). Not to mention that there will be a finite and probably short list of question/answer combos. Most of the variation will be in the questions I imagine and it will be in the form of noise (synonyms and different sentence structures).

They make it very easy to scrape too. The question text is plain text (no need to OCR that) and you can run that through your favorite NLP package to extract useful structure to work from. The answers are each displayed as individual images on the page, making it trivial to know how many possible answers there are. That also makes it easier to run your favorite OCR package on the individually sliced images.

Always glad to see people cracking away at this problem though. It's great fodder for machine learning enthusiasts.

001sky2• 14 years ago

Trolling is the art of feigning one or more of the following: empathy, ignorance, sincerity.

-- They are, in this way, like terrorists. The more you let them in, the greater the damage.

001sky2• 14 years ago

There are all kinds of uses for social engineering.

We should not be surprsised not all of the are pleasant.

Nor should we be surprised, that some are more (or: less) transparent than others.

Afal• 14 years ago

To be honest I don't think this will stop trolls, or at least not the ones that actually believe what they say. There will be still some trolls that will fill in the correct word in the captcha and say stupid shit in the comment. Asking for someone to think morally for a brief second is not going to stop people. This is "just another captcha" to most.

I don't think it's effective against bots either. If a bot can read the three options it has a 1/3 chance of getting it right (assuming that the bot would choose randomly). In fact from the sample captcha it looks like one of the words is not like the others and therefore you could possibly eliminate that making it a 50:50 chance (again assuming it picks it randomly, that there's an odd one out, and that the odd one out is unrelated to the text above).

Although, I suppose that the idea of the captcha is to show awareness, rather than actually stop spam or trolling. It's neat, but not a great spamkiller.

pbhjpbhj2• 14 years ago

>or at least not the ones that actually believe what they say //

I thought the point of trolling was to say things [that you didn't necessarily believe] simply to provoke an emotional or forceful response from the reader.

If you believe it you're not trolling; possibly flame-baiting?

fchollet• 14 years ago

Not a serious attempt at implementing the concept. They only seem to have a few different questions in total. About 6.

Also even with a serious question DB, it would be extremely easy to crack, because of the small pool of answer words and the fact that it's always the negative one you're supposed to pick. With such a low entropy problem, some basic ML and it would be over.

lmkg• 14 years ago

My first reaction to this is "that's dumb." But then, I remember a few studies I've heard of. Basically, if people are made to write an essay in favor of a position (like for or against abortion), even if it's one they initially disagree with, their opinions will be measurably more sympathetic to that position afterwards. So maybe this isn't entirely dumb after all.

Obviously, there's a big gap between writing an essay and typing a single word. Supposedly, the trigger was having to consider the reasoning behind a position, which you don't have to do to type just one word. But who knows, maybe the cumulative effect could actually do something.

Note also that this isn't a real captcha, in the traditional sense. It doesn't work by keeping the trolls out. It works by making them slightly less troll-ish, through repeated exposure. While outwardly similar, it's a very different model, and you would not expect any immediate effect.

xyzzy123• 14 years ago

Right, the term for this is psychological priming.

e.g: http://livinglifewithoutanet.wordpress.com/2010/08/17/honest...

I think it's interesting that you could A/B test this; prime only some users and see if it reduces trolling in that group.

bengl3rt• 14 years ago

What about sociopaths?

sethrin• 14 years ago

Your post advocates a

(x) technical ( ) legislative ( ) market-based ( ) vigilante

approach to fighting spam. Your idea will not work. Here is why it won't work. (One or more of the following may apply to your particular idea, and it may have other flaws which vary from state to state.)

( ) Spammers can easily use it to harvest email addresses ( ) Mailing lists and other legitimate email uses would be affected ( ) No one will be able to find the guy or collect the money ( ) It is defenseless against brute force attacks (x) It will stop spam for two weeks and then we'll be stuck with it ( ) Users of email will not put up with it ( ) Microsoft will not put up with it ( ) The police will not put up with it ( ) Requires too much cooperation from spammers (x) Requires cooperation from too many of your friends and is counterintuitive ( ) Requires immediate total cooperation from everybody at once ( ) Many email users cannot afford to lose business or alienate potential employers ( ) Spammers don't care about invalid addresses in their lists ( ) Anyone could anonymously destroy anyone else's career or business ( ) Ideas similar to yours are easy to come up with, yet none have ever worked ( ) Other:

Specifically, your plan fails to account for

( ) Laws expressly prohibiting it ( ) Lack of centrally controlling authority for email ( ) Open relays in foreign countries ( ) Ease of searching tiny alphanumeric address space of all email addresses (x) Asshats ( ) Jurisdictional problems ( ) Unpopularity of weird new taxes ( ) Public reluctance to accept weird new forms of money ( ) Huge existing software investment in SMTP ( ) Susceptibility of protocols other than SMTP to attack ( ) Willingness of users to install OS patches received by email ( ) Armies of worm riddled broadband-connected Windows boxes ( ) Eternal arms race involved in all filtering approaches ( ) Extreme profitability of spam ( ) Joe jobs and/or identity theft ( ) Technically illiterate politicians ( ) Extreme stupidity on the part of people who do business with spammers (x) Dishonesty on the part of spammers themselves ( ) Bandwidth costs that are unaffected by client filtering ( ) Outlook ( ) Other:

and the following philosophical objections may also apply:

( ) Any scheme based on opt-out is unacceptable ( ) SMTP headers should not be the subject of legislation ( ) Blacklists suck ( ) Whitelists suck ( ) We should be able to talk about Viagra without being censored ( ) Countermeasures cannot involve wire fraud or credit card fraud ( ) Countermeasures cannot involve sabotage of public networks ( ) Sending email should be free ( ) Why should we have to trust you and your servers? ( ) Incompatiblity with open source or open source licenses (x) Feel-good measures do nothing to solve the problem ( ) Temporary/one-time email addresses are cumbersome ( ) I don't want the government reading my email ( ) Killing them that way is not slow and painful enough ( ) Other:

Furthermore, this is what I think about you:

(x) Nice try, dude, but I don't think it will work. ( ) This is a stupid idea, and you're a stupid person for suggesting it. ( ) Nice try, assh0le! I'm going to find out where you live and burn your house down!

jrockway• 14 years ago

I like these, but it doesn't really work this time.

The problem is that their idea won't work. "emotions" is too small a search space. A spam bot could just answer "terrible" to every question and still send a good amount of spam.

reCaptcha works because there are a lot of words in English and a lot of source material that needs to be digitized. That makes it infesible to guess one word, let alone two words at the same time.

This emot-i-captcha requires a human to come up with an unbounded number of unique questions and answers, which is impossible.

dlss• 14 years ago

I think the bigger problem is that the questions are currently in plain text, which, since they require a human to write and once solved are solved forever, makes the approach as implemented a sort of quixotic reverse captcha.

If they switch to image based questions, the potential problems for human solvers are magnified.

anthonyb• 14 years ago

No it doesn't.

Choose three words, one each from a group of positive (I like), neutral and negative (I don't like), along with something from a big list of news items which are good or bad.

Now they have to choose (I like + good news) or (I don't like + bad news). Adding new words or new news items are independent of each other, and it doesn't seem that hard compared to traditional captchas.

pjscott• 14 years ago

That template used to be very popular -- pessimism feels cool, doesn't it? -- but it lost a lot of credibility when spam filters got really good. I hardly ever get spam anymore.

thaumasiotes2• 14 years ago

As a consequence of living an atypical life, I have a somewhat dimmer view of spam filters. I don't get spam, but I also don't get genuine email from people I've actually corresponded with in the past. Seriously... if I maintain correspondence with a small set of qq numbers, and another qq number emails me, giving the exact same name as a different email account which I've written back and forth to... I shouldn't even see this new message, because it came from qq? This is the state of the art in gmail?

And Gandi's webmail service won't even file mail from China as spam... it won't deliver it at all.

Evbn2• 14 years ago

It didn't lose credibility, it shot down every technique except the one that worked, Bayesian filtering. Plus I guess a handful of web mail services taking over hundreds of millions of accounts, which may have not been foreseen 10 years ago.

bitwize• 14 years ago

Oh look, a Voight-Kampff test for the interbutts.

Fun fact about people with no empathy: they are often profoundly skilled at feigning empathy when they feel they are being tested for it.

jawns• 14 years ago

Apparently, that's one of the marks of a psychopath:

http://www.reddit.com/r/IAmA/comments/otl8z/iama_diagnosed_p...

pbhjpbhj2• 14 years ago

Post hoc ergo propter hoc ?

nc17• 14 years ago

I can see this idea used to keep out people who don't conform to groupthink. You could ask questions about abortion, god, taxes, etc.

The problem is that the questions need to be very clever to avoid giving away the site's bias (and thus the answers).

wamatt• 14 years ago

>"How does that make you feel"

Mu.

Obviously it's dead simple to guess the 'correct' answer.

But wrapping it in a faux test of one's humanness is an off-putting premise, that can come across morally condescending (which probably wasn't the intention).

While I support civil rights and the group behind it is probably well intentioned, it doesn't follow automatically that this instance of execution is a good idea.

Also, it's probably fallacious to assume only 3 emotions (or any at all for that matter), are occurring in the user.

"meh", "bored" or "mildly annoyed" etc, are perfectly valid answers too, and does not mean we are jerks for feeling that way :p

jamestc• 14 years ago

I agree. It's a pretty shallow idea. Is it a terrible thing to be "fascinated" by something that is supposedly terrible? Fascination doesn't actually necessitate a polarized reaction. And, really, exclusionary tactics/gatekeeping with opinions? These seem like awfully short-sighted solutions to a deeper problem.

And, what's more, trolls aren't hard to mobilize if provoked. Start implementing this and I wouldn't doubt seeing a small movement devoted to going the extra mile to offend people with "humane" sensibilities.

leddt• 14 years ago

In addition to the questionable morals of such a system, I think it is technically flawed.

Most captchas work because the words they present to users are mostly random assortments of letters. This one, however, presents real dictionary words. It is pretty conceivable that a bot, armed with a dictionary file, would have a pretty high rate of success correctly identifying at least one of the three words.

If it then randomly chooses one of the identified words, it would have a rate of success at solving the captcha nearing 33%. This is extremely high when your bot is trying to solve thousands of captchas per minute.

desireco42• 14 years ago

This whole idea is repulsive to me, chairman Mao would probably welcome it, but that is all. It is dumb idea made by people who are not interested in solving captcha, but pushing their own vision on how morals should be. If in the age of internets we can't stand people with differing opinions and all have to be beaten into same mold, then we are really screwed.

politician• 14 years ago

Is it ethical to use this form of mind control? Who decides how you should be reprogrammed?

mooism2• 14 years ago

Oh come now. This is hardly less ethical than the methods used by e.g. Fox News or the Daily Mail, but we seem to be mostly comfortable with that.

KMag• 14 years ago

"This enables a simpler and more effective way of keeping sites spam free as well as taking a stand for human rights."

Over-stating the security implications of their system for the purposes of getting press for an upcoming event is short-sighted.

gogetter• 14 years ago

How would Steve Jobs have fared on these empathy tests?

What if they asked questions about workers in Chinese factories?

D'oh!

LolWolf• 14 years ago

Obligatory xkcd reference: http://xkcd.com/233/

Anyways, point is, although it's a good idea, I don't see it being either wide-spread or significantly more effective than re-CAPTCHA, et al., nor do I see it 'aiding' in emotion, so to speak.

But, I would love for someone to prove me wrong on this, because it is an interesting approach.

Fend Off Trolls, Bots and Jerks With ‘Empathy’ Test

Related Stories