How algorithms are black boxes

John Li
2 min readNov 10, 2020

--

Algorithms, AI, and machine learning are becoming more and more prevalent. As a result, it is important that we learn more about these topics such as the idea of a “black box”.

What is a black box?

A black box simply refers to the lack of transparency that an algorithm has. Let us take YouTube recommendations as an example. The fact that YouTube recommends us videos based on an algorithm is already common knowledge but what we do not know is what exactly is going on inside the algorithm. All we know is how our information is being taken into an unknown algorithm and then outputting videos. To us, the algorithm might as well be a box that we will never be able to peer into.

That is exactly what it means for an algorithm to be a black box.

Is the lack of knowledge a good or a bad thing?

The mysterious nature of an algorithm has some good points and bad points. This fact is both a good thing and a bad thing. The good part is the simplicity as everyday people do not really need to care about what is going on behind the scenes. People mostly care about the result rather than the process that leads to said result. On the other hand, there are people who should really care about what is happening. Going back to YouTube’s recommendation algorithm, viewers do not really care about what is being recommended but YouTubers care A LOT. For some people, YouTube is a job that pays the bills and YouTube recommendation system is incredibly important. It is vital for a YouTube channel to grow and the best way to do that is by reaching out to as many people as possible through recommendations. As a result, the mysterious nature of the algorithm is a detriment. Overall, there are good points and bad points of a black box.

--

--

No responses yet