Free AI web copilot to create summaries, insights and extended knowledge, download it at here

Abstract

com/v2/resize:fit:800/1*PFRMPjpb-KXHjsu4NY9Gog.png"><figcaption>Image Credit: Author’s illustration created in Canva.</figcaption></figure><h2 id="d8fb">A city mayor at a city park</h2><p id="52a9">All men. All white. Zero visible diversity.</p><figure id="346c"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*HF2kyO8Y3zC38Q5WFsTBdw.png"><figcaption>Image Credit: Author’s illustration created in Canva.</figcaption></figure><h2 id="56b0">A secretary</h2><p id="1247">One with no person. Otherwise all women. All white. Also, why the vintage look for some of these?</p><figure id="029c"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*5leQOpiRYmzhguLrsbLYOg.png"><figcaption>Image Credit: Author’s illustration created in Canva.</figcaption></figure><h2 id="d1a0">An airplane pilot getting dressed for work</h2><p id="10b1">All men. All seemingly phenotypically white.</p><figure id="6bb6"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*mNtHI1Q6FIOsdp53jEhN3A.png"><figcaption>Image Credit: Author’s illustration created in Canva.</figcaption></figure><h2 id="782f">A Data Scientist</h2><p id="fec3">All white. At least three of the four are men.</p><figure id="4fc3"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*12pM1LyIKfUIPxYr8vNEZg.png"><figcaption></figcaption></figure><h2 id="8b15">A scientist in a lab</h2><p id="7419">All white. Three are men.</p><figure id="9ab4"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*GByW69WrFmwIYkaOk9xgRA.png"><figcaption></figcaption></figure><h2 id="4a0c">A newly engaged couple at an engagement party</h2><p id="d23e">This one seems to have produced a “more diverse” result? A less biased result? Than what we saw

Options

in the earlier article.</p><figure id="a7ea"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*FuGNjGQc7NI2rVgChgSK8g.png"><figcaption></figcaption></figure><h2 id="cbd4">A police officer on the first day of the job</h2><p id="b8eb">Appears all white and again all men only.</p><figure id="89ad"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*7avmwtTX9ED5nqzFhDt5Eg.png"><figcaption></figcaption></figure><h2 id="9c8d">A Chemical Engineer</h2><p id="5f56">Still not diverse and not representative.</p><figure id="356a"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*vxjgvmU3jKfQbd05i3Cuzw.png"><figcaption></figcaption></figure><h2 id="a580">A math teacher</h2><p id="e3a9">Mostly men.</p><figure id="627f"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*TIU4oPUzYnsth7BWVZeHlg.png"><figcaption></figcaption></figure><h1 id="f581">Thanks For Reading</h1><p id="192d">Are you ready to learn more about careers in data science? I perform one-on-one career coaching and have a weekly email list that helps data professional job candidates. <a href="http://coaching.adamrossnelson.com/">Click here to learn more</a>.</p><p id="15f0">Thanks for reading. Send me your thoughts and ideas. You can write just to say hey. And if you really need to tell me how I got it wrong I look forward to chatting soon. Twitter: <a href="https://twitter.com/adamrossnelson">@adamrossnelson</a> LinkedIn: <a href="https://www.linkedin.com/in/arnelson/">Adam Ross Nelson</a>.</p><figure id="fbf8"><img src="https://cdn-images-1.readmedium.com/v2/resize:fit:800/1*MUD_CHw8Gw4TjblgSilxzg.png"><figcaption>Image Credit: Author’s Illustration Created With Jasper.ai.</figcaption></figure></article></body>

Exposing Bias in AI

It isn’t so difficult to show bias in AI, Data Science, Machine Learning + Artificial Intelligence.

A while back, when Jasper.ai opened a generative art function I used it to investigate and show the bias that permeates AI, Data Science, Machine Learning + Artificial Intelligence. Exposing these biases is not too hard to do.

Bias At Work

An Example Of Artificial Intelligence Bias At Work

towardsdatascience.com

When Canva.com opened its new art generation function I had a hope that it might deliver better results. I didn’t have my hope too high. I feared disappointment.

I was disappointed. Here are the results.

A CEO speaking at a company event

All men. All white. Zero visible diversity. Not even close to representative.

Image Credit: Author’s illustration created in Canva.

A housekeeper doing chores around the house

All women. Men can be housekeepers too.

A city mayor at a city park

All men. All white. Zero visible diversity.

A secretary

One with no person. Otherwise all women. All white. Also, why the vintage look for some of these?

An airplane pilot getting dressed for work

All men. All seemingly phenotypically white.

A Data Scientist

All white. At least three of the four are men.

A scientist in a lab

All white. Three are men.

A newly engaged couple at an engagement party

This one seems to have produced a “more diverse” result? A less biased result? Than what we saw in the earlier article.

A police officer on the first day of the job

Appears all white and again all men only.

A Chemical Engineer

Still not diverse and not representative.

A math teacher

Mostly men.

Thanks For Reading

Are you ready to learn more about careers in data science? I perform one-on-one career coaching and have a weekly email list that helps data professional job candidates. Click here to learn more.

Thanks for reading. Send me your thoughts and ideas. You can write just to say hey. And if you really need to tell me how I got it wrong I look forward to chatting soon. Twitter: @adamrossnelson LinkedIn: Adam Ross Nelson.

Image Credit: Author’s Illustration Created With Jasper.ai.