Racism and bias in artificial intelligence: a sad truth of our technological advancement

Racism has been a prevalent issue in modern times. Though humans have improved relations between the different races, many are wondering whether artificial intelligence will do the same mistake. 

Artificial intelligence has been progressing as years pass by, advancing in areas humans never imagined they could reach. In the modern times, artificial intelligence have been a huge help for the population, performing menial and difficult tasks ahead. But to do such tasks, artificial intelligence must be taught from someone. Someone can program thoughts, ideas into machines. 


Parth here, we wanted to tell you more about DataCamp, where you can master the foundations of AI, Python and Machine Learning from scratch by creating a relevant portfolio of projects that you can add to your resume.

Whether you’re a complete beginner or want to take your career to the next level, the courses on DataCamp have provided me with strong foundations in Data science, Artificial Intelligence and Programming and started me on a brand new career path. 

728x90.jpg

These courses are for:

  1. Anyone with zero programming experience, who wants to become a Data Science and Machine Learning Expert.

  2. Anyone looking for a course to teach you everything about ML and Data Science and learn the latest tools and technologies used at Netflix, Google and Amazon.


Unknown bias towards certain races could also be programmed. 

A sad truth of the technological industry is that it mostly led by rich white men. One prime example of this the case of ten large technological companies in Silicon Valley —that dominated the technological industry— did not employ a black woman of colour. Three of them had no black employees at all. If AI is programmed with such men with no diversity involved, bias and prejudice is bound to happen. 

The fragments of racism in artificial intelligence was shown in a video in 2017that displayed a soap dispenser that would only release soap into white hands. This error occurred because the company, Technical Concept, have not allowed a person of colour to test the product against coloured skin in the development team, causing the machines unable to detect colour.

Another study that proved the idea of racism in artificial intelligence is the study conducted in March 2018 where it was shown that cars driven by AI are more likely to drive into black pedestrians. The same issue that occurred with Technical Concept occurs here, the driverless cars fail to recognise coloured skin. 

Machines are blank-slate until someone programs it. Source: towardsdatasceince.com

Machines are blank-slate until someone programs it. Source: towardsdatasceince.com

What makes this is even more fretting is that artificial intelligence is often involved in employment, meaning they choose who to employ and why. This becomes an issue because they are also involved in law enforcement and if such biases is prevalent, then discrimination in employment can become a larger problem (Poverty, unemployment and crime rising in minority communities). There’s also a possible future where artificial intelligence will take part in hospital admission, school exclusions and hiring processes. 

An expert on this issue is Mike Bugembe have said that ‘the progress of AI have a larger effect on the lives of black people that goes far beyond who does or doesn’t get to use hand soap’ He believes that perpetuating racism will have a long-lasting harm to minorities, therefore it is crucial that technological companies hire more black and minority people to allow machines to remove those unconscious biases. 

‘It’s no secret that within six years, pretty much 98% of human consumer transactions will go through machines. And if these machines don’t see us, minorities, then everything will be affected for us. Everything.’ Mike replies  

If humanity has chosen to allow machines to be a part of their life, it is important that machines have no bias and can allow to see the world with no prejudice involved. To ensure this is possible, technologial compaines must allow their development team to have more diversity than what they have now.

Thumbail: TechCrunch