All Data are Bias

Because Humans are Bias!

OK, calling humans are bias sounds very negative, perhaps let me paraphrase it, “Humans have preferences.”

“Wait, how so?” Let’s go to one of our favorite moments, stopping at the ice cream parlor. This is what goes through in your mind. You will automatically filter flavors that you know you will not enjoy! First-level preferences! Moving on, you will then be given all the different factors like price, weather, and ratings, to narrow down to flavors that can maximize your utility with the cost you are willing to pay. There you go, you have selected the “best” combination of ice cream flavors, happy as can be as you go into summer/or hot weather here in Singapore.

You can see from here, all humans have preferences! With the data collected, the ice cream parlor may proceed to start analyzing the data to make business decisions on.

Extend that further, the ice cream parlor definitely serves a particular demographic given its location, accessibility, pricing, and many more, serving mostly a particular group of customers.

I hope to keep the newsletter open to as many as possible. If you have benefitted much from the newsletter and will like to join my cause, consider making a “book" donation. Link at the bottom of the newsletter or here. :)

We have to accept the fact that data will be biased, that is the point I want to make. The earlier we accept the fact that data are biased, especially those generated with humans involved, we can start working on processes that can either tease out these bias or rectify the harm or damage caused quickly and effectively.

Machine learning thrives on preferences and differences! Think about that! In other words, in a world of homogeneity, machine learning will never work well, will never dig out those hidden rules or structures in our data (and perhaps we should even conclude and say there is no Free Will).

What are your thoughts? Do share them in the comment. :)

Recommended Content

If you are interested in alternative views on building Artificial Intelligence, I recommend Jeff Hawkin’s book. Start with “On Intelligence”. :)

If you still remember Garry Kasparov playing against Deep Blue and want to get his perspectives during and after the game, check out his book “Deep Thinking”.