Artificial Intelligence (AI) is rapidly transforming the world we live in. Even though we are not aware of it, there are already plentiful applications around us targeted at making our lives relaxed with this type of technology, a development that will without a doubt grow in upcoming years.
With this rapid growing and enthusiastic growth of AI, we should not neglect the ethical side of its exploitation and in particular, the biases that may be part of these future applications.
But what are biases? In our daily lives we make many decisions, from the brand of beer we drink, car we drive, the movies we watch, the books we choose to read… and most of the time, without even knowing it, those decisions are biased.
In people’s case, it is our brain that picks the answer and majority of the time it is based on our likings and experience, knowledge or views, which are accountable for that biased thinking. But what happens when these choices are made by machines?
With the advancement in technology, the arrival of Big Data and mainly A.I. , more and more, it is these systems that make judgements. Those judgements sometimes may be inappropriate, but much more often intelligent systems are starting to be able to determine whether I will be able to clear the exam, granted credit or have my credit card limit increased.
Within AI, it is in machine learning where these biases can appear, as it is built on historical information. As far as machine learning is involved, biases can occur in various parts of this process and not only in the data, as it would seem logical to think.
First of all, we must clearly specify which business related problem we want to solve. Think of a company that wants to do a recruiting process to find the best worker. This person could be the one with the highest income, the one who advances the fastest in the company or the one who is the most knowledgeable. This first step, which many times we do not take into consideration when valuing bias, it has to be accurate enough.
The second step is the data we are going to choose. It may be that they are not enough demonstrative for the issue we want to solve or that they already include their own bias. An example could be Amazon’s algorithm, which only proposed men for certain positions because their training data contained a majority of male staff.
Finally, it is very important how you choose the information to build the algorithm, ie what variables will be relevant in each case. If we take the example of the selection of employees we could choose the age, work experience, education, merits and so on. The exactness of the model will differ on the choice of characteristics and it will not always be easy to understand the implicit bias included in these data.
Given the significance of biases as they relate to Machine Learning, the last five years have seen an rise in the number of articles, research and tools to ease them.
On the research side, various related scientific papers and many other AI development companies are starting to build tools to measure these biases and find ways to explain the decisions made by their algorithms.
Google already has algorithms in trail to enhance bias analysis and explain ability.
It is crucial that the people engaged in the growth of these algorithms, as well as the companies that use them, ask themselves if they are taking into account these biases and if they are able to justify the decisions made by their algorithms. While it is true that most technology brands have issued their ethical codes, and at the European level there are references with 7 major points Trustworthy AI, there is still long time to go in terms of ethics
While it is true that most large tech companies have published their ethical codes
A recent Capgemini report on ethics in organizations revealed, among other results, that companies that took ethics into account had a 44-point advantage with the NPS (Net Promoter Score), the most generally used index to measure customer loyalty and their affinity to suggest that company or that product to others. It is good news that ethics in companies can be a aim to increase buyers and therefore sales, so more and more companies will be interested in this section.
Thanks to the digital revolution, now more than ever clients have much more impact on companies, being able to be prescribers or detractors in the social networks of products or companies. Let’s use that power so that the progression of AI does not leave ethics aside and thus helps us all.