So. Many. Notifications. — #DiversityRecognition

Diversity Gone Viral

5 million selfies, and counting…

UPDATE — 2:47 EST, June 21, 2017: The app is back

As the majority of us have either been brought up with or adopted a beautiful “free love” attitude toward culture mixing, and super easy access to our ancestry details abound — we have become completely obsessed with ethnicity. Our own ethnicity, and the ethnicity of others.

Which is why we weren’t exactly shocked when our Diversity and Ethnicity app went viral.

In fact, over 5 million curious people (at the time of writing) have uploaded their images to our app, anxious to know what ethnic tale the algorithms would tell!

Here’s an example of how we show you your ethnicity makeup from a simple selfie — Try it now

The feedback has been humbling

User response has ranged from amazement and praise, to displeasure and offense. And we totally understand why.

While most users will get a spot-on result, we acknowledge that the ethnicity classifiers currently offered (Black, White, Asian, Hispanic, ‘Other’) fall short of representing the richly diverse and rapidly evolving tapestry of culture and race.

Determining, with confidence, the ethnicity(s) by which we identify is a tall order, even for sophisticated algorithms — as the connection between people and their heritage is so deeply personal and complex. Yet, through the viral success of our Ethnicity and Diversity app, we have had an amazing, eye opening lesson in the taxonomy of ethnicity.

“Hi, there!
Let me start by saying that your mission is admirable and impressive!
Although, I do wish that your recognition system was capable of identifying South & South-East Asian features.
I have seen this problem multiple times where my friends of Indian descent, and myself (being mixed with both African and Indian ancestry) where there are situations where your analyzer is unable to detect these non-East Asian (not Chinese, Japanese, Korean, etc.) features.
Please look into fixing this, as it will help in correctly identifying people’s backgrounds. All the best!”

— Just one of the myriad comments we have recieved the past few days.

We’ve learned that many people of Latin descent are turned off by the term “Hispanic”, and prefer Latinx. And that lots of mixed folks identify as HAPA, a Hawaiian word for “half”, “part”, or “mixed”. We were also interested to know that our Native American friends are identifying as Amerindian.

These are just a few examples of the ways these groups self identify, and maybe more importantly, the way they want others to identify them.

Change is coming…

We are so grateful for the unprecedented amount of ‘real-world’ feedback we’ve received about our Ethnicity and Diversity app — as it will serve as the cornerstone for ongoing research, and guide us in improving the inclusive integrity of future versions of our classifiers.

We are now, more than ever, conscious of the magnitude of our challenge in recognizing the complexities of people, and realizing our responsibility to teach our technology that ethnic classification is more than multidimensional — It’s personal.

Read more about #DiversityRecognition, and find out what diversity means to us at Kairos:

And here’s what it looks like to ‘go viral’:

Kairos’ mission is to make it easy for any business to benefit from face analysis, enriching the experience between humans and machines, and being the premier partner for anything to do with facial recognition.