I am going to object to the idea that "Tech" can be racist, it can be functional or nonfunctional, the people who use it can be good or bad the tech is agnostic

I am going to object to the idea that "Tech" can be racist, it can be functional or nonfunctional, the people who use it can be good or bad the tech is agnostic. Now this argument is made more interesting in the age of ML and AWS Rekognition in particular, but my understanding is that ML decisions/classifications are made based on training data, if the model had been trained with different data it would've produced different results. Ultimately this ML is just math run at a massive scale with a ton of data points, it isn't intended to show how the world should be rather express according to the math how it actually works, the math the model is objective it doesn't care about feelings or bias or prejudice it just classifies images according to an algorithm based on training data. Can someone help me understand the counterpoint that technology is inherently ist? I agree it can be used to push ist agendas or perpetuate causes of *ism but technology itself seems like it would be pretty objective and if it is producing uncomfortable or bad results based on it's math either the tech is bad, or it is trying to show us something we may not like.