Stories
Slash Boxes
Comments

SoylentNews is people

posted by janrinok on Wednesday July 01 2015, @05:29AM   Printer-friendly
from the did-you-really-just-call-her-that? dept.

Google Photo tries to categorize your pictures automatically. Until very recently, it had a failure mode in which its classification for some pictures of humans was "Gorillas".

Google reacted [and apologised] very quickly when they got a complaint from a black woman who had been misclassified.

When Brooklyn-based computer programmer Jacky Alcine looked over a set of images that he had uploaded to Google Photos on Sunday, he found that the service had attempted to classify them according to their contents. Google offers this capability as a selling point of its service, boasting that it lets you, “Search by what you remember about a photo, no description needed.” In Alcine’s case, many of those labels were basically accurate: A photograph of an airplane wing had been filed under “Airplanes,” one of two tall buildings under “Skyscrapers,” and so on.

Then there was a picture of Alcine and a friend. They’re both black. And Google had labeled the photo “Gorillas.” On investigation, Alcine found that many more photographs of the pair—and nothing else—had been placed under this literally dehumanizing rubric.

Speculating, it's possible that their software is heavy on statistical matching and it's really hard to debug, which is why they wound up simply deleting "Gorilla" from the list of possible categories.

http://www.slate.com/blogs/future_tense/2015/06/30/google_s_image_recognition_software_returns_some_surprisingly_racist_results.html


Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 4, Insightful) by penguinoid on Wednesday July 01 2015, @05:35AM

    by penguinoid (5331) on Wednesday July 01 2015, @05:35AM (#203634)

    The 800 lb gorilla is that some people are easily offended.

    --
    RIP Slashdot. Killed by greedy bastards.
    • (Score: 1) by Absolutely.Geek on Wednesday July 01 2015, @11:15AM

      by Absolutely.Geek (5328) on Wednesday July 01 2015, @11:15AM (#203706)

      Yes and it is because we are living in the future; algorithms can now be racist. Now where the hell is my flying car?

      --
      Don't trust the police or the government - Shihad: My mind's sedate.
    • (Score: 5, Informative) by jimshatt on Wednesday July 01 2015, @12:11PM

      by jimshatt (978) on Wednesday July 01 2015, @12:11PM (#203725) Journal
      To be fair, she was explicitly NOT easily offended and reacted in a VERY understanding and reasonable manner.
      • (Score: 3, Informative) by Anonymous Coward on Wednesday July 01 2015, @12:23PM

        by Anonymous Coward on Wednesday July 01 2015, @12:23PM (#203733)

        > To be fair, she was explicitly NOT easily offended and reacted in a VERY understanding and reasonable manner.

        Why ruin a good case of confirmation bias with facts?

        The irony is that penguinoid was too easily offended by the story and so made up his own narrative to justify his reaction. That seems to be de rigueur for his fellow travelers.

      • (Score: 2) by penguinoid on Wednesday July 01 2015, @05:21PM

        by penguinoid (5331) on Wednesday July 01 2015, @05:21PM (#203842)

        Yes but a lot of her friends seemed to be much more offended. And I suspect a lot of her probably-not-friends like Jesse Jackson will be going totally ape-shit.

        --
        RIP Slashdot. Killed by greedy bastards.
      • (Score: 2) by tynin on Wednesday July 01 2015, @11:30PM

        by tynin (2013) on Wednesday July 01 2015, @11:30PM (#204018) Journal

        Indeed. When metro.co.uk reached out to them to use the picture, and they responded, "sure, feel free under the CC-BY-ND license."

        Very reasonable license choice! They were ultra cool through the whole event.

    • (Score: 5, Insightful) by sjames on Wednesday July 01 2015, @12:45PM

      by sjames (2882) on Wednesday July 01 2015, @12:45PM (#203738) Journal

      RTFA! In spite of Google accidentally making a comment that has long been used as a hateful racial slur, the mis-identified person rather matter of factly pointed out the obvious error without going into any sort of rant.

      How did you manage to translate that into "easily offended"?

    • (Score: 4, Touché) by VortexCortex on Wednesday July 01 2015, @07:42PM

      by VortexCortex (4067) on Wednesday July 01 2015, @07:42PM (#203934)

      The 800 lb gorilla is that some people are easily offended.

      "800 Lbs", so now you're fat shaming?!

  • (Score: 2) by MichaelDavidCrawford on Wednesday July 01 2015, @05:40AM

    by MichaelDavidCrawford (2339) Subscriber Badge <mdcrawford@gmail.com> on Wednesday July 01 2015, @05:40AM (#203635) Homepage Journal

    The mac os word 5 spellchecker flagged childcare, suggested kidnaper as a replacement.

    --
    Yes I Have No Bananas. [gofundme.com]
    • (Score: 2) by Kromagv0 on Wednesday July 01 2015, @01:58PM

      by Kromagv0 (1825) on Wednesday July 01 2015, @01:58PM (#203758) Homepage

      I'll drink to that.
      Word 97

      --
      T-Shirts and bumper stickers [zazzle.com] to offend someone
    • (Score: 2) by MozeeToby on Wednesday July 01 2015, @06:47PM

      by MozeeToby (1118) on Wednesday July 01 2015, @06:47PM (#203891)

      Up until very recently, Chrome's spell check would flag Apatosaurus and suggest Brontosaurus. I always assumed it was a joke by some engineer whose favorite dinosaur was "stolen" from him.

  • (Score: 3, Insightful) by GoonDu on Wednesday July 01 2015, @05:48AM

    by GoonDu (2623) on Wednesday July 01 2015, @05:48AM (#203636)

    What if the person in the photograph do look like a gorilla? On a serious note, if an image recognition algorithm misclassifies, I would not go as far as to cry 'racism'. Just simply a mistake.

    • (Score: 1) by Pino P on Wednesday July 01 2015, @01:53PM

      by Pino P (4721) on Wednesday July 01 2015, @01:53PM (#203756) Journal

      What if the person in the photograph do look like a gorilla?

      Break ties with the pose. The hip of a gorilla and the hip of a human lead to very different postures even if both are walking bipedally. If it looks like a gorilla and walks like a gorilla, it's a gorilla. If it looks like a human and walks like a human, it's a human. If it looks like a gorilla and walks like a human, it's a well-known hoax known as Bigfoot. In other words, a human.

      • (Score: 2) by physicsmajor on Wednesday July 01 2015, @04:56PM

        by physicsmajor (1471) on Wednesday July 01 2015, @04:56PM (#203827)

        Nice ideas that don't work at all with deep learning. These neutral nets have no human prior information other than labels on their training data. If you want it to learn to differentiate two things it's misclassifying, you have to feed it more correctly labeled training data on either side of the error.

      • (Score: 2) by vux984 on Wednesday July 01 2015, @07:19PM

        by vux984 (5045) on Wednesday July 01 2015, @07:19PM (#203915)

        Break ties with the pose.

        Not sure how that's going to work with shots that don't show completely show them; what do you do if the 'posture' isnt in the shot?
        Unless you mean, "if it looks like a selfie angle assume human"? That would work; or better still just tag such photos "mouth breather".

  • (Score: 5, Insightful) by m2o2r2g2 on Wednesday July 01 2015, @05:49AM

    by m2o2r2g2 (3673) on Wednesday July 01 2015, @05:49AM (#203637)

    Now I know this will come off as insulting, and I really don't want it to be, but...

    *TO A COMPUTER* Gorillas have the same humanoid shape as people, and unfortunately the colour of the hair covering on most Gorillas are black. So a person who has darker skin is more likely to fit into that bucket in a primitive image recognition algorithm (matches shape and colour).

    This is not Google trying to be racist/insulting, this is just something that highlights how hard that feature is to implement correctly. The people need to take a chill pill and realise this was not a deliberate attack on them. It was a machine that made a mistake. There was no malice involved (as management would not allow the extra engineering time to program malice in).

    We are going to run into a lot of these problems as people try to project emotions and intentions onto machines.

    • (Score: 3, Interesting) by Ethanol-fueled on Wednesday July 01 2015, @06:08AM

      by Ethanol-fueled (2792) on Wednesday July 01 2015, @06:08AM (#203641) Homepage

      This kind of thing is nothing new. Years ago HP was scolded for their face-tracking technology not working with Black people. [gizmodo.com]

      Although I can't help but chuckle, because I'm an asshole, this isn't a symptom of racism but rather shoddy design that isn't tested and developed using enough real-life cases. This is also the kind of problem that occurs when you offload your beta-testing onto the public rather than do it yourself.

      Somewhat related, there's a lot of fun to be had with Google search's auto-suggestions. For example this, [torontosun.com] this, [aurumahmad.com] and this. [lancs.ac.uk]

       

      • (Score: 4, Interesting) by zocalo on Wednesday July 01 2015, @07:31AM

        by zocalo (302) on Wednesday July 01 2015, @07:31AM (#203651)
        There was also the issue with a camera's facial recognition tech that had a habit of assuming that Asian people were blinking or even had their eyes closed and helpfully suggesting that the photographer retake the photo, compounded by the fact that the camera vendor was Asian. There are so many subtle nuances as to what might be in an image that's it's not surprising that classification errata like this happen from time to time, but sometimes you do wonder just how large and varied the test data sets were.
        --
        UNIX? They're not even circumcised! Savages!
        • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @12:56PM

          by Anonymous Coward on Wednesday July 01 2015, @12:56PM (#203740)

          compounded by the fact that the camera vendor was Asian.

          Probably they outsourced the software development to a cheap American sweatshop. ;-)

    • (Score: 2) by Gravis on Wednesday July 01 2015, @06:17AM

      by Gravis (4596) on Wednesday July 01 2015, @06:17AM (#203642)

      *TO A COMPUTER* Gorillas have the same humanoid shape as people

      oh SUUUUURE. what are you going to claim next, a computer labeling a donkey as a horse is understandable?!

      • (Score: 2) by arslan on Wednesday July 01 2015, @06:40AM

        by arslan (3462) on Wednesday July 01 2015, @06:40AM (#203645)

        The donkey would be quite happy...

        • (Score: 2) by c0lo on Wednesday July 01 2015, @06:52AM

          by c0lo (156) Subscriber Badge on Wednesday July 01 2015, @06:52AM (#203647) Journal
          The horse would be offended to be tagged an ass, a wild ass [wikipedia.org] even more.
          Understandable; wouldn't you feel the same?
          --
          https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
        • (Score: 5, Funny) by fritsd on Wednesday July 01 2015, @09:19AM

          by fritsd (4586) on Wednesday July 01 2015, @09:19AM (#203686) Journal

          Did you hear that, Shrek? I'm a *STEED* !!1!

    • (Score: 5, Interesting) by VLM on Wednesday July 01 2015, @12:04PM

      by VLM (445) on Wednesday July 01 2015, @12:04PM (#203720)

      *TO A COMPUTER* Gorillas have the same humanoid shape as people

      Its actually worse, not just basic shape. There's a classic pix we've all seen but I can't find a link from one of the "save the gorillas in the mist" or whatever fundraising campaigns where they played photo games with cropping and angles and the gorilla looks like a human with a punched nose. You know, to get sympathy and financial donations you're going to use a pix that looks kinda human for PR, not a pix of them taking a dump or something. Not for all gorillas, not all pix of that one gorilla, but there is at least one pix of one gorilla one time, out there with VERY human ratios of eye separation to mouth width, eye separation to eye/mouth distance, gorilla eyes are basically human eyes, human looking mouth, basically a little plastic surgery, some make up, and a touch of photoshopping and that gorilla from that cropped specific angle could pass for a "people of walmart" entrant. It wouldn't take much work. You need the hair just right and the sun just right at just the right angle so you don't notice the brow ridge issues and the nose isn't so prominent, etc etc. The TLDR point I'm trying to make is there exist a subset of pictures of gorilla faces that pass quite well for possibly very bad pictures of human faces.

      Or turning it around given the right blackface makeup and really weird lighting conditions to get things just right and give me a weird facial expression to make and do my hair just right, and a very carefully contrived and cropped pix of my head could easily pass for a pix of a somewhat deformed gorilla.

      Its like those people that torture their pets by making them wear crazy animal Halloween costumes, no google, my coworkers dog really isn't Thomas the Tank Engine its just a dog wearing the craziest costume you've ever seen under ideal camera conditions. Those people must feed their dogs 50 pound of jerky to get them to tolerate that stuff, dogs are even worse than toddlers WRT that kind of thing.

      I suspect a huge amount of the butthurt is creationist / non-evolution types trying to stealthily proclaim they are completely unrelated to lower primates, which is obviously pretty stupid. I don't care how much you claim it offends Jesus, at least some gorillas look like humans and vice versa because "Duh!" we're related, admittedly very distantly. You can act as offended as you want about photographs but the molecular biologists pretty much don't give a shit, we're closely related end of story. I suspect there's -kin people problems too, "well I feel I was born to be a vampire squid and it emotionally offends me when people say I'm biochemically a primate all that matters is how I feel so you have to treat me as if my delusions are real, you must love my tentacles".

      Combine the above paragraphs and you got people who win "dog looks like its owner" contests, so considering we're biologically about a bazillion times closer related to gorillas than Shetland Sheepdogs or parrots, there should logically be about a bazillion times more "gorillas and dudes look alike" pix, for factual reasons that are really going to piss off creationist types.

      • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @02:18PM

        by Anonymous Coward on Wednesday July 01 2015, @02:18PM (#203772)
        Well that escalated quickly...
    • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @12:31PM

      by Anonymous Coward on Wednesday July 01 2015, @12:31PM (#203736)

      > There was no malice involved (as management would not allow the extra engineering time to program malice in).

      Of course there was no malice involved. It is the unconscious assumptions of the developers that get programmed in because management does not allow the extra engineering time to even consider other possibilities. That's something you get for free when your developers aren't a monoculture.

    • (Score: 4, Insightful) by tempest on Wednesday July 01 2015, @01:19PM

      by tempest (3050) on Wednesday July 01 2015, @01:19PM (#203744)

      software is heavy on statistical matching

      I can see how a computer could mistake that for a gorilla, however if we're leveraging statistics I'd say this says more about a poor implementation. There are more pictures, by a few orders of magnitude of black people than of gorillas, so I'd expect the software should be mistaking gorillas for black people not the other way around.

    • (Score: 2) by Runaway1956 on Wednesday July 01 2015, @02:48PM

      by Runaway1956 (2926) Subscriber Badge on Wednesday July 01 2015, @02:48PM (#203783) Journal

      "*TO A COMPUTER* Gorillas have the same humanoid shape as people,"

      Uhhhh - how about to people as well? Understand that I have never had "normal" vision. Although there are circumstances in which I can see some things better, and quicker than people around me, there are many more circumstances in which I'm the last person to see what everyone else sees. I have been confused by pics of monkeys and apes plenty of times. It depends on how much of the animal or the person you can see, and the posture. Quality of color, contrast, and focus contributes to the potential for confusion.

      It's possible that I didn't spend enough time studying my mama's face when I was an infant, but I've seen a fair number of photographs that have confused me, however briefly.

    • (Score: 2) by bzipitidoo on Wednesday July 01 2015, @02:59PM

      by bzipitidoo (4388) on Wednesday July 01 2015, @02:59PM (#203789) Journal

      Knew the racist angle would be immediately trumpeted. But where are all the offended Creationists? Fact is, all of us look similar to chimps and gorillas for the very good reason that we are closely related. We look much more like chimps than we look like, say, dogs or cats.

      • (Score: 2) by Freeman on Wednesday July 01 2015, @07:06PM

        by Freeman (732) on Wednesday July 01 2015, @07:06PM (#203903) Journal

        You don't hear creationists trumpeting in, because we don't care about how you think we look like Gorillas. Why would it be an aberration for God to make an animal that has some features resembling humans? You can think we evolved from apes, but I choose to disagree. There's not much of a discussion here, thanks.

        --
        Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
    • (Score: 2, Insightful) by albert on Wednesday July 01 2015, @05:02PM

      by albert (276) on Wednesday July 01 2015, @05:02PM (#203830)

      The comparison wouldn't be offensive if it didn't have some truth to it. We'd just laugh if the label was "starfish" or "ice cream".

      Suppose you are comparing noses. Compare one of those wide flat African-style noses with one of those skinny long European-style noses. Which is more likely to be mislabeled as a gorilla nose?

      Suppose you are comparing brow ridges. The same applies. Add in the brow ridge of an Australian Aborigine though, and they win the gorilla look-alike contest by far.

      Suppose you are comparing eye color. Lots of European natives get a special exemption from the gorilla category.

      Suppose you are comparing side face profiles, which wasn't applicable in this case, or slight ridges (sagittal keel) running along the top of the head. The gorilla's lower face projects a bit, the forehead slopes back a bit, and there is that ridge. These features are all somewhat common in Africans, but relatively uncommon in others.

      • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @10:21PM

        by Anonymous Coward on Wednesday July 01 2015, @10:21PM (#203984)

        > The comparison wouldn't be offensive if it didn't have some truth to it.

        Are you honestly ignorant of the long history of racial slurs against black people comparing them to apes and gorillas?

        Here's one against Michelle Obama. [wistv.com]

        • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @10:40PM

          by Anonymous Coward on Wednesday July 01 2015, @10:40PM (#203995)

          "Are you honestly ignorant of the long history of racial slurs"

          And, are you honestly ignorant of the fact that people resemble monkeys and apes? Are you honestly ignorant that coloration tends to reinforce the resemblances? You might address the issue, rather than slamming reasonable observations. No one gives a fuck about your sensitive little feelings.

          • (Score: 0) by Anonymous Coward on Thursday July 02 2015, @12:06AM

            by Anonymous Coward on Thursday July 02 2015, @12:06AM (#204023)

            > And, are you honestly ignorant of the fact that people resemble monkeys and apes?

            Which has absolutely nothing to do with why the comparison is 'offensive.'

    • (Score: 2) by Beryllium Sphere (r) on Thursday July 02 2015, @04:44AM

      by Beryllium Sphere (r) (5062) on Thursday July 02 2015, @04:44AM (#204095)

      Better testing might have caught this before it stepped on someone's toes, so an accusation of neglect might fly.

  • (Score: 3, Funny) by penguinoid on Wednesday July 01 2015, @05:50AM

    by penguinoid (5331) on Wednesday July 01 2015, @05:50AM (#203638)

    And I'm willing to bet that the software was written by an almost exclusively white team of programmers.*

    *Displays his knowledge that there are very few black programmers in the most offensive way possible.

    --
    RIP Slashdot. Killed by greedy bastards.
    • (Score: 3, Interesting) by penguinoid on Wednesday July 01 2015, @06:22AM

      by penguinoid (5331) on Wednesday July 01 2015, @06:22AM (#203643)

      Hm, I thought I could trick it by uploading some pictures of albino gorillas (I edited and blurred the pictures so it can't cheat too much). It correctly identified them as animals. I'm impressed -- image recognition is hard.

      --
      RIP Slashdot. Killed by greedy bastards.
      • (Score: 2) by Thexalon on Wednesday July 01 2015, @01:58PM

        by Thexalon (636) on Wednesday July 01 2015, @01:58PM (#203759)

        Have you tried using chimpanzees instead of albino gorillas? Looking at this resemblance [seedsofdoubt.com] between a prominent political figure and a chimpanzee seems like that might work to fool it.

        --
        The only thing that stops a bad guy with a compiler is a good guy with a compiler.
        • (Score: 2) by Kromagv0 on Wednesday July 01 2015, @02:16PM

          by Kromagv0 (1825) on Wednesday July 01 2015, @02:16PM (#203767) Homepage

          I think I might need to upload that picture to facebook and see what it's facial tagging does. I know who is going to be labeled a chimp and who is going to be labeled George W. Bush if I can tag them. Now I want to do the same for this one as well [nobeliefs.com].

          --
          T-Shirts and bumper stickers [zazzle.com] to offend someone
      • (Score: 1, Informative) by Anonymous Coward on Wednesday July 01 2015, @08:22PM

        by Anonymous Coward on Wednesday July 01 2015, @08:22PM (#203951)

        I remember when Snowflake made the cover of National Geographic in 1967.

        At first, the Barcelona Zoo was not aware that Snowflake was a unique specimen. They sent a message to Sabater Pi saying, "Please send more white gorillas."
        [...]
        Snowflake fathered twenty-two offspring by three different mates, or "dams". Six of his offspring survived to adulthood. None of Snowflake's offspring were albino, but all should be heterozygous, recessive carriers, for the albino gene. His grandchildren have 50% mathematical probability for carrying the albino gene. If both parents were albino gene carriers they have 25% chance of producing an albino offspring.

        Snowflake (gorilla) [wikipedia.org]

        -- gewg_

  • (Score: 2) by takyon on Wednesday July 01 2015, @05:52AM

    by takyon (881) <takyonNO@SPAMsoylentnews.org> on Wednesday July 01 2015, @05:52AM (#203639) Journal
    --
    [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
  • (Score: 4, Informative) by Gravis on Wednesday July 01 2015, @06:44AM

    by Gravis (4596) on Wednesday July 01 2015, @06:44AM (#203646)

    image in question: https://pbs.twimg.com/media/CIoW7wBWoAEqQRP.png [twimg.com]
    actual gorilla: https://c1.staticflickr.com/7/6169/6151464249_1824cde43d_b.jpg [staticflickr.com]

    a very dark face with a wide nose and seemingly surrounded by black. is it really a surprise it was labeled as being gorillas?

    • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @07:29AM

      by Anonymous Coward on Wednesday July 01 2015, @07:29AM (#203650)

      The human brain treats faces separately from general recognition tasks. Maybe Google should implement something similar to its algorithms. Have some "looks like a face" detector and divert anything that qualifies to an extra algorithm that is specialized in recognizing human faces.

      • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @07:34AM

        by Anonymous Coward on Wednesday July 01 2015, @07:34AM (#203652)

        Have some "looks like a face" detector and divert anything that qualifies to an extra algorithm that is specialized in recognizing human faeces.

        Same shit will happen.

      • (Score: 4, Informative) by linuxrocks123 on Wednesday July 01 2015, @07:44AM

        by linuxrocks123 (2557) on Wednesday July 01 2015, @07:44AM (#203654) Journal

        A face detector wouldn't be enough, because gorillas have faces. Looking at the photos, it's a selfie with a black girl's face taking up most of the picture; the guy is way in the background.

        The girl has long black hair on both sides of her face, and her body (which would have been helpful for disambiguation) isn't in the photo. I don't have much experience with image recognition software, but I think it's likely her hair combined with the photo being a closeup of just her face is probably what did it.

        Probably the ultimate fix will be changing the heuristics to err way, way on the side of labeling gorillas people, rather than the other way around. If a real gorilla photo ends up in the "people" category, the gorilla won't be offended.

        • (Score: 2) by linuxrocks123 on Wednesday July 01 2015, @07:47AM

          by linuxrocks123 (2557) on Wednesday July 01 2015, @07:47AM (#203655) Journal

          Oh, also, I would guess it's highly likely Google already has a face detector. That's one of the easiest ways to find people in a photo.

        • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @09:09AM

          by Anonymous Coward on Wednesday July 01 2015, @09:09AM (#203683)

          A face detector wouldn't be enough

          If you had read my comment instead of just skimming through it, you would have noticed that I didn't claim it would be. All the face detector would do it to identify the areas needing special attention. The real work would have to be in the specialized routines that are to be run on the areas of special attention.

          • (Score: 2) by linuxrocks123 on Wednesday July 01 2015, @09:21AM

            by linuxrocks123 (2557) on Wednesday July 01 2015, @09:21AM (#203688) Journal

            Apologies. That's a good idea: detect faces, and then spend extra effort trying to classify the species of the face if it's not obvious from other parts of the photo.

            Many photos wouldn't need the extra effort, even if there's a face, because you know from the wearing of clothing and/or shape of the body that it is or isn't a human.

            It would be nice if Google would post some information about their algorithms in general, how they went wrong in this case, and what they're doing to fix it.

        • (Score: 1) by esperto123 on Wednesday July 01 2015, @12:19PM

          by esperto123 (4303) on Wednesday July 01 2015, @12:19PM (#203731)

          I think having gorillas been classified as people would also make people lose their shit (no pun intended), although probably not as much.

          But it got me thinking, we can differentiate between gorillas and humans quite easily, even though we are VERY similar, what do we use to differentiate? face proportions? wrinkles? bone structure?

          • (Score: 2) by Freeman on Wednesday July 01 2015, @07:12PM

            by Freeman (732) on Wednesday July 01 2015, @07:12PM (#203908) Journal

            Intelligence.

            --
            Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
          • (Score: 2) by tangomargarine on Wednesday July 01 2015, @08:23PM

            by tangomargarine (667) on Wednesday July 01 2015, @08:23PM (#203952)

            I think having gorillas been classified as people would also make people lose their shit (no pun intended),

            What pun?

            --
            "Is that really true?" "I just spent the last hour telling you to think for yourself! Didn't you hear anything I said?"
    • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @04:07PM

      by Anonymous Coward on Wednesday July 01 2015, @04:07PM (#203809)

      Don't start throwing the racist card at this... But I've noticed some peoples faces do look like Gorillas, AND... I've heard young children say "mommy, look at the monkey" when seeing a black person. It was totally innocent, and no racism was implied.

      • (Score: 2, Touché) by albert on Wednesday July 01 2015, @05:16PM

        by albert (276) on Wednesday July 01 2015, @05:16PM (#203839)

        We learn to suppress such thoughts as we grow older. We learn there are things you can't say or even think.

        Other good ones involve obese and pregnant people being confused, especially when male. :-)

        • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @07:45PM

          by Anonymous Coward on Wednesday July 01 2015, @07:45PM (#203937)

          Yes. My point was... If innocent children can make that mistake, then so can a computer algorithm. I have yet to see a racist program, unless it was intentional. Good point about the mistaken pregnancy label, done it myself.

          • (Score: 0) by Anonymous Coward on Thursday July 02 2015, @02:39AM

            by Anonymous Coward on Thursday July 02 2015, @02:39AM (#204055)

            > I have yet to see a racist program, unless it was intentional.

            That is circular reasoning - software can't be racist because, by your definition, it is ignorant so therefore you've never seen an unintentionally racist program.

            And yet here we have an example of one.

        • (Score: 0) by Anonymous Coward on Thursday July 02 2015, @02:44AM

          by Anonymous Coward on Thursday July 02 2015, @02:44AM (#204057)

          > We learn to suppress such thoughts as we grow older. We learn there are things you can't say or even think.

          Maybe you learned that, but you learned the wrong lesson.

          What normal people learn is that being needlessly insulting is a shitty way to live your life.

    • (Score: 1) by jpkunst on Sunday July 05 2015, @11:56AM

      by jpkunst (2310) on Sunday July 05 2015, @11:56AM (#205266)

      Yes. Look at the lips. Black people have thick lips, gorillas (and other apes) have thin lips.

  • (Score: 1, Funny) by malloc_free on Wednesday July 01 2015, @08:11AM

    by malloc_free (3034) on Wednesday July 01 2015, @08:11AM (#203665) Journal

    ...and she was correctly identified as a 'crack whore'. Something is obviously working XD

    • (Score: 2) by zeigerpuppy on Wednesday July 01 2015, @10:38AM

      by zeigerpuppy (1298) on Wednesday July 01 2015, @10:38AM (#203699)

      I got "Nubian princess" on my images,
      Gonna say google got it right for once.

      • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @02:21PM

        by Anonymous Coward on Wednesday July 01 2015, @02:21PM (#203774)
        What's a Nubian? /banky
    • (Score: -1, Flamebait) by Anonymous Coward on Wednesday July 01 2015, @11:09AM

      by Anonymous Coward on Wednesday July 01 2015, @11:09AM (#203703)

      I was just giving soylent a second chance after being turned off by the juvenile troll shit that was taking hold here.

      This shit isn't funny, and fuck you.

      I'll check back in a few months and see if the community has grown up at all.

      (guess what race i am...)

      • (Score: 1) by deathlyslow on Wednesday July 01 2015, @12:00PM

        by deathlyslow (2818) <wmasmith@gmail.com> on Wednesday July 01 2015, @12:00PM (#203718)

        Try setting your thresholds higher.

      • (Score: 1, Flamebait) by Runaway1956 on Wednesday July 01 2015, @02:07PM

        by Runaway1956 (2926) Subscriber Badge on Wednesday July 01 2015, @02:07PM (#203764) Journal

        FFS, no one gives a damn what race you are. There are some pretty serious, informative posts here, and there are some juvenile bullshit posts here as well. I don't care WHAT race you claim to be, if you can't see the obvious, then you're dumber than a box of rocks.

        Don't like the stupid bullshit? Change your browsing settings. That, or grow a pair. That right to never be offended? That one is made-up bullshit. You don't HAVE any such right.

        Speaking of being offended - some of us are offended at your whining and sniveling.

      • (Score: 1, Funny) by Anonymous Coward on Wednesday July 01 2015, @02:24PM

        by Anonymous Coward on Wednesday July 01 2015, @02:24PM (#203776)

        (guess what race i am...)

        ...crack whore?

      • (Score: 1) by malloc_free on Thursday July 02 2015, @12:17AM

        by malloc_free (3034) on Thursday July 02 2015, @12:17AM (#204025) Journal

        Ah ok was supposed to be a juvenile joke (kinda going for a Southpark thing), and at least I had the decency to not post as myself, not as AC. If you are to chicken shit to do the same, I for one will not lament your departure.

        • (Score: 0) by Anonymous Coward on Thursday July 02 2015, @02:41AM

          by Anonymous Coward on Thursday July 02 2015, @02:41AM (#204056)

          Jokes like that reveal more about the teller than intended, mr non-anonymous "malloc_free."

  • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @12:18PM

    by Anonymous Coward on Wednesday July 01 2015, @12:18PM (#203730)

    Deja vu all over again

    Microsoft Encarta got in trouble for labelling a family as "monkey", the monkey bars in the playground behind them was the reason given
    http://www.zdnet.com/article/microsoft-sued-for-racist-application/ [zdnet.com]

  • (Score: 2) by bziman on Wednesday July 01 2015, @01:46PM

    by bziman (3577) on Wednesday July 01 2015, @01:46PM (#203753)

    The error is probably in the training data. In photos, 90% of gorillas are very dark in coloring. In photos, only about 10% of humans are very dark in coloring. The algorithm probably thinks that coloring is the primary differentiator between humans and gorillas.

    The problem is that the training data is pulled off the web, rather than manually curated. In order to get enough data for this kind of AI, you basically have to crowd source the training set. But then you bake in these sorts of biases.

    Perhaps for the next iteration they should include equal numbers of each race of human and each subspecies of primate? It would take much longer and produce generally poorer results for the wider audience, but it might reduce outliers.

    • (Score: 2) by Beryllium Sphere (r) on Thursday July 02 2015, @04:40AM

      by Beryllium Sphere (r) (5062) on Thursday July 02 2015, @04:40AM (#204092)

      A CMU AI graduate I know pointed out that if, hypothetically, Google had used the web at large for training data then it would have been poisoned by all the racist web sites out there.

  • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @02:17PM

    by Anonymous Coward on Wednesday July 01 2015, @02:17PM (#203769)

    Has a good chance at a class action lawsuit for theft of identity.

  • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @02:24PM

    by Anonymous Coward on Wednesday July 01 2015, @02:24PM (#203777)

    This is the result of Social Justice. Everything is horrible offensive, and everything must have a hate campaign against it.

    • (Score: 0) by Anonymous Coward on Wednesday July 01 2015, @04:20PM

      by Anonymous Coward on Wednesday July 01 2015, @04:20PM (#203815)

      Yup. The "Dukes of Hazard General Lee" Hot Wheels car is now offensive and banned, but now selling for up to $300 on eBay.

    • (Score: 4, Informative) by eof on Wednesday July 01 2015, @05:02PM

      by eof (5559) on Wednesday July 01 2015, @05:02PM (#203831)

      Did you read the article? The guy pointed it out and went on with his life.