That’s what this piece argues: Of course technology perpetuates racism. It was designed that way. | MIT Technology Review.
I disagree. Technology sometimes perpetuates racism, but it is often due to the fact it is NOT designed to account for racism. Sometimes machine learning software cannot perceive non-white faces correctly because they are trained with only white faces and cannot account for non-white faces. Sometimes search engines result in racist results, based on racist queries. AI systems can be made racist by engaging with a multitude of racists. If you feed systematic racist data based on redlining into your banking system or prison data into your justice software system, then those systems will make racist decisions. In all these cases, the fact that the systems are not designed to account for racism (or sexism or any form of discrimintation) is the problem. They need to be designed to account for these things.
Only when technology is designed to account for racism will it stop perpetuating racism.