{"id":339296,"date":"2017-04-26T06:29:37","date_gmt":"2017-04-25T22:29:37","guid":{"rendered":"http:\/\/techcrunch.com\/?p=1481383"},"modified":"2017-04-26T06:29:37","modified_gmt":"2017-04-25T22:29:37","slug":"faceapp-apologises-for-building-a-racist-ai","status":"publish","type":"post","link":"https:\/\/people.utm.my\/asmawisham\/faceapp-apologises-for-building-a-racist-ai\/","title":{"rendered":"FaceApp apologises for building a racist AI"},"content":{"rendered":"

<\/p>\n

If only all algorithmic bias were as easy to spot as this: FaceApp, a photo editing app that uses a neural network for editing selfies<\/a>\u00a0in a photorealistic way\u00a0has apologized for building a racist algorithm.<\/p>\n

The app lets users upload a selfie or a photo of a face, and offers\u00a0a series of filters that can then be applied to the image to subtly or radically alter its\u00a0appearance \u2014 its appearance-shifting effects include\u00a0aging and even changing gender.<\/p>\n

The problem is\u00a0the app also included a so-called \u201chotness\u201d filter, and this\u00a0filter was racist. As\u00a0users pointed out<\/a>\u00a0the filter\u00a0was lightening skin tones to achieve\u00a0its mooted\u00a0\u2018beautifying\u2019 effect. You can see the filter\u00a0pictured above in a before and after shot of President Obama.<\/p>\n

In an emailed statement apologizing for the racist algorithm, FaceApp\u2019s\u00a0founder and CEO\u00a0Yaroslav Goncharov told us:\u00a0\u201cWe are deeply sorry for this unquestionably serious issue. It is an unfortunate side-effect of the underlying neural network caused by the training set bias, not intended behaviour. To mitigate the issue, we have renamed the effect to exclude any positive connotation associated with it. We are also working on the complete fix that should arrive soon.\u201d<\/p>\n

As the Guardian<\/a> noted earlier, the app has had a surge in popularity in recent weeks \u2014 perhaps prompting FaceApp to realize the filter\u00a0had a problem.<\/p>\n

FaceApp has temporarily changed the name of the offending filter from \u201chotness\u201d to \u201cspark\u201d,\u00a0although it would have been smarter\u00a0to pull it from the app entirely until a\u00a0non-racist replacement was\u00a0ready to ship. Possibly they\u2019re being distracted by the app\u2019s moment of viral popularity (it\u2019s apparently adding around 700,000 users per day).<\/p>\n

While the underlying AI tech powering FaceApp\u2019s effects includes code from some open source libraries, such as Google\u2019s Tensorflow, Goncharov confirmed to us that the data set used to train the \u201chotness\u201d\u00a0filter is\u00a0its own, not a public data set. So there\u2019s no getting away from where the blame lies here.<\/p>\n

Frankly it would be hard to come up with a better (visual) example of the risks of biases being embedded within algorithms. An machine learning model\u00a0is only as good as the data it\u2019s fed \u2014 and in FaceApp\u2019s case the Moscow-based team\u00a0clearly did not train their algorithm on a diverse enough dataset. We can at least thank them\u00a0for illustrating the lurking\u00a0problem of\u00a0underlying algorithmic bias in such a visually impactful way.<\/p>\n

With AI being handed control of more and more systems there\u2019s a pressing need for algorithm accountability<\/a> to be fully interrogated, and for robust systems to be put developed\u00a0to avoid embedding human biases into our machines. Autonomous tech\u00a0does not mean \u2018immune to human flaws\u2019, and any developer that tries to claim otherwise is trying to sell a lie.<\/p>\n","protected":false},"excerpt":{"rendered":"

\"\" If only all algorithmic bias were as easy to spot as this: FaceApp, a photo editing app that uses a neural network for editing selfies in a photorealistic way has apologized for building a racist algorithm. Read More<\/a><\/p>\n

\n<\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a> <\/img><\/a>\n<\/div>\n

\"\"\/ <\/p>\n","protected":false},"author":5817,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"categories":[25],"tags":[59,66,26],"class_list":["post-339296","post","type-post","status-publish","format-standard","hentry","category-technology","tag-media","tag-techcrunch","tag-technology"],"_links":{"self":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts\/339296","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/users\/5817"}],"replies":[{"embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/comments?post=339296"}],"version-history":[{"count":0,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/posts\/339296\/revisions"}],"wp:attachment":[{"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/media?parent=339296"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/categories?post=339296"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/people.utm.my\/asmawisham\/wp-json\/wp\/v2\/tags?post=339296"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}