Andrea Uličná (Scholarship Runner-Up) - Free Will vs Recommender Systems

Join over 2 million students who advanced their careers with 365 Data Science. Learn from instructors who have worked at Meta, Spotify, Google, IKEA, Netflix, and Coca-Cola and master Python, SQL, Excel, machine learning, data analysis, AI fundamentals, and more.

Start for Free
The 365 Team 17 Aug 2021 12 min read

Andrea Uličná won second prize in the 365 Scholarship competition. Her essay stood out from the rest with its original structure, intriguing points, and a rich variety of real-life arguments. Andrea is a student at the Lutheran High School Tisovec, Slovakia and we wish her great success in her studies.

We believe you'll enjoy her piece as much as we did.

*Formatting and images added by the 365 team

The illusion of choice: Is the recommendation algorithm taking away your free will?

computer screen with icons of platforms that use recommendation algorithms

Disclaimer: This essay was written based on the research done by searching tools using recommendation algorithms.

Searched term: recommendation algorithm

Some people for sure know what a recommendation algorithm is. Yet, others might think that it is a magic formula from Harry Potter. So let’s set it straight and ask our friend Google.

A recommendation algorithm ”also known as a recommender system, is software that analyzes available data to make suggestions for something that a website user might be interested in, such as a book, a video or a job, among other possibilities.” (, 2019) A recommendation algorithm is, therefore, a tool able to browse accessible data and according to a user’s preferences adjust the results to match them with user’s interests.

Searched term: history of recommendation algorithms

Though being associated mainly with e-commerce sites, such as Amazon or Netflix, Recommender systems powered by algorithms were born before these companies. To be exact, in 1992. The very first recommender system was called Tapestry. It was developed at the Xerox Palo Alto Research Center and what an innovation it was! Allowing people overwhelmed with hundreds of thousands of documents sent via email to define their own personal filter for those they actually want to receive. In addition to this function, Tapestry could as well handle collaborative filtering, meaning that “people collaborated to help people perform filtering by recording their reactions to the documents they read.” (Goldberg et al., 2019) Later on, GroupLens extended the basic idea brought about by the Xerox Palo Alto Research Center and set the basis for further development. GroupLens developers adapted recommender systems to filter within larger sets of information. (Goldberg et al., 2019)

Searched term: the present use of recommendation algorithms

Nowadays, almost every company, which wants to be competitive enough has its own recommendation algorithm incorporated into their websites. Take the example of Airbnb, Netflix, Amazon, Instagram or Facebook. Each one suggesting the best places or products. The same applies to smaller companies which are dominant in one specific region, like book or property selling sites within a country.

Having discussed the term recommendation algorithm, the history of the system as well as its role in today’s society, it is clear what algorithms are, where they are from and what do we use them for. Most probably, an average person would be satisfied and have already marked the task of “understanding recommendation algorithms” as done on his to-do-list. On the other hand, the person with a curious nature driven by the awareness that there is always something new to learn continues reading, being eager to find our more interesting facts which are yet to come.

Searched term: effects of recommendation algorithms on our lives

Getting done more in a shorter time

Since the original purpose of the recommendation algorithms was to solve the work-related problem, the invention benefited both, the employer and also employees. Employees are no more getting lost in the endless lists of emails, invoices, products or any other type of data used within companies. At the same, employers enjoy higher productivity from people working for them, more often than not resulting in better customer service leading to higher profits. And the virtuous circle goes on, bringing more money to be distributed among the staff. (Goldberg et al., 2019)

Case study: Getting the results you do not except but need (personal experience)

I have a personal experience of recommendation algorithm benefiting my life and future career. Back then, I did not really know how they worked, yet, I got results I did not except but very much needed.

Back in the summer of 2018, I was looking for a summer internship. Having been a part of an educational program, in which we established a school company with my classmates, the previous academic year, I wanted to do something business related. Any guesses what a high school student do if he or she wants to get a summer internship? Hits the Google search. It is just as natural as breathing.

Mainly thanks to that search I am now working on my own business idea dedicated to children self-development. But how did I get to start something like that?

As I was a proactive student in all sorts of extracurricular activities throughout primary and also secondary school, the systems behind algorithms knew that I had been searching business competitions for high school student, workshops focused on presentational or communicational skills or calls for Slovak participants in European projects before this internship search. That is why my results included also a call for participants in a mentoring program devoted to the development of business ideas. I did not really hesitate to join that mentoring program and today I could proudly say that my idea has developed into quite a successful project.

Would you expect a recommendation algorithm to give you such a boost towards your future career? I definitely would not. But I am not the only one to experience a success story like this. My friend got the opportunity to travel to Mexico for a week-long JA Global Youth Forum about leadership, as she was searching for a summer opportunity abroad and saw an application page in her results, once again produced by the recommendation algorithm.

Based on these two stories and many more presented in relevant online media, I think that in some cases recommendation algorithms are able to use collected data about a person in a way that provides information that could change one’s life for the better.

Searched term: effects of recommendation algorithms on our lives

Switch from the original purpose

However, with the improvements in recommendation algorithms, the power of these systems has reached a new level. Being present almost everywhere, they affect and change far broader aspects of life.

Trapped in a filter bubble

A filter bubble is defined as „a state of intellectual isolation that allegedly can result from personalized searches when a website algorithm selectively guesses what information a user would like to see based on information about the user, such as location, past click-behavior and search history.” (Wikipedia, 2019) Basically, recommendation algorithms create a bubble for each and every user looking up information, which presents the cultural as well as ideological viewpoints and beliefs of the user. This is done by the information the user provides the algorithm with.

Case study: Are we losing the ability to think critically enough?

Though filter bubbles have certain advantages, in my opinion, there is more to fear and praise. Let’s take the recent presidential election in Slovakia as an example.

According to research done by a platform Youth Council of Slovakia, which represents all of the youth clubs and non-governmental organizations in Slovakia, 45% of young Slovaks got information about presidential election form social media on daily basis, while 51% of them occasionally. (Rada mládeže Slovenska, 2019)

The presidential campaign on social media prior election was quite massive. Therefore, the majority of the young people were a part of the target group politicians were trying to address. And this is when recommendation algorithms showed their power. Most of the young people had already had a profile or two on social media, and so algorithms used to filter posts already knew what the certain user sympathize with.

What was the result? If you previously followed and liked the posts of a candidate representing one point of view, it was very unlikely that you would see a post from a candidate with different opinions. (Rada mládeže Slovenska, 2019)

Some people argued that there is nothing wrong with having your social feed full of the posts of the side you are most likely to vote for. But does not that mean that you do not get the opportunity to truly evaluate the best possible candidate? Whilst it might be true that you previously liked the ideas of another candidate or political party, your opinion might have changed based on the different circumstances which life brings. Therefore one’s ability to think critically gets weakened.

A quite easy way how to get the proper information is to look for them in more relevant sources than social media. Yet, most of the people do not get out of their comfort zone to question things which appear in their social feed, as they presume them as an only option available. Thought the impact recommendation algorithms might have on the result of presidential or any other election is still being discussed, the doubt whether or not could be this tool used to support propaganda-like filter bubbles is there as well.

The political election is only one example of the possible threats of the recommendation algorithms for our freedom of choice and autonomy. Very recently, there was a book promoting false theory about Republicans and Democrats in the USA being featured in the top selling books on Amazon. Amazon declined to answer any questions about the book’s position in page’s lists, but the majority of trustworthy media, including NBC News, do agree that the book contains false information. (Collins, 2019)

As the position of books on Amazon is generated algorithmically, this case once again shows the negative outcomes of using recommendation algorithms.

Searched term: all aspects of recommendation algorithms

All in all, living in the modern area when we are able to access any information from any place, it is upon us to ensure we truly have the rights we proclaim to have, without enabling technology (developed by ourselves) to take it away from us at any time and in any way. Therefore, I believe it is important to be aware of the strengths, weaknesses, opportunities and also threats of recommendation systems. Following four paragraphs sum up the content of the essay discussed above.


Originally, the idea of a tool for filtering huge amounts of data was inspired by a problem experienced by many people. Tapestry was born in order to save time, in other words, to make work more effective. Therefore the predominant strength and advantage of recommendation algorithms is the effectiveness it brings. Moreover, algorithms have been and are boosting changes in the IT sector. So, their abilities will always be needed in the technical fields aspiring advancements. (Maruti Techlabs, 2019)


Like nothing, recommendation algorithms are not perfect. The performance of the recommendation algorithms depends on the available data. It would not be able to suggest people anything relevant without enough data about them, their hobbies and interests. Moreover, we would not be people if we did not change our mind on an everyday basis. This is something recommendation algorithms simply cannot handle. In this case, they lack the most important characteristic natural to humans – the ability to think critically. Therefore, changing one’s mind confuses them, letting them unable to provide a relevant result of a search. (Lam, Frankowski and Riedl, 2019)


By many, recommendation algorithms are seen as an irreplaceable ingredient for business to be successful in the future. Once computers improved enough to be able to run more complex algorithms, two fields – Machine Learning and Deep Learning started flourishing. What is more, these actions allowed Artificial Intelligence to come at the scene and literally change the whole game of business success. Hence, recommendation algorithms will for sure play a huge role in the world of future business. (Ismail, 2019)


The fact that disadvantage recommendation algorithms is that these systems are overexposed to security threats. It does not matter whether it is an interest of companies running them or of the competition. They are very likely to be the victims of a security breach, making them, once again, irrelevant or even dangerous for people. (Lam, Frankowski and Riedl, 2019)


Collins, B. (2019). On Amazon, a Qanon conspiracy book climbs the charts — with an algorithmic push. [online] NBC News. Available at: [Accessed 28 Apr. 2019].

Goldberg, D., Nichols, D., Oki, B. and Terry, D. (2019). [PDF] Using collaborative filtering to weave an information tapestry | Scinapse | Academic search engine for paper. [online] Scinapse. Available at: [Accessed 28 Apr. 2019].

Ismail, S. (2019). Why Algorithms Are The Future Of Business Success. [online] Available at: [Accessed 28 Apr. 2019].

Lam, S., Frankowski, D. and Riedl, J. (2019). Do You Trust Your Recommendations? An Exploration of Security and Privacy Issues in Recommender Systems. [online] Available at: [Accessed 28 Apr. 2019].

Maruti Techlabs (2019). How do Recommendation Engines work? And What are the Benefits?. [online] Maruti Techlabs. Available at: [Accessed 28 Apr. 2019].

Rada mládeže Slovenska (2019). Mladých vo voľbách ovplyvnil prevažne internet. [online] Available at: [Accessed 28 Apr. 2019]. (2019). What is recommendation engine? - Definition from [online] Available at: [Accessed 28 Apr. 2019]. Wikipedia (2019).

Filter bubble. [online] Available at: [Accessed 28 Apr. 2019].  

The 365 Team

The 365 Data Science team creates expert publications and learning resources on a wide range of topics, helping aspiring professionals improve their domain knowledge, acquire new skills, and make the first successful steps in their data science and analytics careers.