https://ift.tt/3z4woHQ Opinion Should we think about the need for some regulations related to the use of data? Photo by Lukas Blazek on...
Opinion
Should we think about the need for some regulations related to the use of data?
In a world in which data is becoming more and more important, should we think about the need for some regulations related to the use of data? In other words: should Data Science be regulated in some way?
In this article, I want to share my thoughts on Data Science and Machine Learning applied to some particular fields and reflect on the possible needs to regulate the use of the data in these fields.
As we all know, the tool is “neutral” in some way; I mean: a tool is a tool; it is not good or bad. It is always the use of the tool that can be good or bad. I mean that Data Science is not good or bad: is the use of the data that can be defined as good or bad. Of course, “good” or “bad” is something related to some personal opinions; so, the question is: there should be some entities which have to define (as law?) that some use of the data is “good” and some other is “bad”?
Let’s see some practical examples…
1) The use of Data Science in Industrial and Manufacturing Engineering
The use of Data Science and Machine Learning in the Industrial field is becoming wider and the applications are increasing. Thinking of the fact that we can predict when precisely to stop a machine to maintenance it, without having delays in the manufacturing processes (the exact maintenance day is predicted), is something like magic.
But I think the troubles always arrive when we have to do with human beings. So, let’s see a particular example.
As a Process and Manufacturing Engineer, one of my daily tasks is to monitor and control the production KPIs; and of course, I do it using python and some Data Science techniques.
What I do is very simple: when a certain product is manufactured, I control if the real production time is consistent with the theoretical production time for this particular product. The fact is that there can be faster and slower operators, but, in the mean, we are just interested if the production times are consistent (otherwise, it means that there are manufacturing issues), regardless of the operators who are working in that particular moment.
But what if in some industry one wants to search for the slower operators to fire them? Yes, of course, there are some privacy issues (at least, here in Italy): one can not present data with operators’ names so that they can be “punished” because they are slow in their work, for example. But of course, sometimes there is no need to present the data with the names; names can just be said outside the meetings…
So, the question is: how can we deal with production data without affecting human factors? Is there a need to regulate Data Science (and the use of the data) in industrial processes? And if yes, how can we do it?
2) The use of data in courts
As we all know, trials are based on facts; and, of course, in courts, facts must be supported by evidence and data.
You have to know that my wife and I love English and to maintain us trained in the use of the language we watch TV series and movies in English. In those days we are watching the TV series called “Bull”. Yes, it is not quite a new one; but, If you don’t know what I’m talking about, the stories have a background on data: doctor Jason Bull (the main character) is a psychologist working as a consultant for lawyers or for people who needs a lawyer. His work is pretty simple: he has to find the best jurors for his clients.
But how does he find the best jurors? Using some Data Science techniques. Practically, he can find data from the jurors (supported by software, in some way) and can define which jurors are the best for his clients.
You know, in courts the important fact is to win and the job of a lawyer is to convince jurors and judges. But what if in the future someone can really choose the best jurors (and the best judges?) based on their behavior, their habits, and their beliefs for their clients? What if life or death sentences will be granted by jurors and judges chosen by a Machine Learning algorithm? Who will win: the algorithm or the human behaviors?
3) Machine Learning and Social Media
I’ve recently read the book “Ten Arguments for Deleting Your Social Media Accounts Right Now “ by Jaron Lanier and it was pretty an incredible reading (I really advise you to read it).
The fact is pretty simple because Machine Learning works in a pretty simple way; the “machine learns” from you, so:
- YouTube advises music (and videos) for you (“you can also like this video”)
- Amazon advises you what to buy (“other people have also bought….”)
- Facebook and the other social media shows you things (news, videos, photos, etc…) based on what you like the most; it means that you can see just a part of the reality.
The Machine Learning algorithms create a model of you so that the social media shows you just what you like: just a part of the reality. This means, for example, that if you have a political idea you will see news consistent with this idea and you will hardly change it; in other words, the software and the algorithm (“the social media”) will reinforce your ideas, your bias, your entire world and you will probably remain the same person, which is basically the contrary of the evolution along a human life.
But what if I’d like to have on YouTube a button like “don’t suggest me other videos: I like different kinds of music so you are not really able to label me”? Well, to “break the algorithm” sometimes I watch videos that are really different from one another (and yes, my dear algorithm: my daughters watch the videos for children with my YouTube account; it is not really me who’s watching!)
In conclusion: what will be the future for human beings related to the use of data? Will human beings be the center of the world in the next future or data (and business?) will be?
If we think about the manufacturing industry we know that automation is replacing humans, and, in my opinion, this can even be good for repetitive tasks…but will Data Science bring the need to use faster robots?
In the end…what will be the future for human beings in a data-driven world?
Will the slowest operators be fired? Will the slowest robots be substituted by faster robots?
Will the jurors in a court be chosen by their behavior by data-driven software so that a party can be the favorite one? Is there a particular need to be labeled by an algorithm as a “rock and roll” lover (I knew it before YouTube told me!)?
I think that the time has come to reflect on these topics and understand if there is the need to regulate the use of the data, in some way and in some contexts.
Some Philosophical Thoughts on Data Science was originally published in Towards Data Science on Medium, where people are continuing the conversation by highlighting and responding to this story.
from Towards Data Science - Medium https://ift.tt/3JpmLZi
via RiYo Analytics
ليست هناك تعليقات