The Era of Algorithms

Algorithms impact day-to-day, but do we know what they are and how they work?

The current age could be called “the age of algorithms” because they control much of the world for better or worse. Machine learning is fast becoming part of everything, improving many existing products, and allowing new products and services to be created. It is the oxygen for the emerging generation of technology.

The algorithm is a sequence of instructions to solve a problem automatically. It is a finite statement with points of input and output, and it executes the instructions systematically. Computerized algorithms have become a key part of the Fourth Industrial Revolution. If the instructions are input with artificial intelligence, machines can learn and improve algorithms.

This technology is used every day to detect diseases, control flights, bank accounts, and even decide which posts to display on social media. Due to their impact on various areas of our lives, it is essential to ask how they work? What criteria do they use?

Algorithmic criteria: the good and the bad

One of the most recurring questions in the age of algorithms concerns is how right and wrong are coded in computers. For this, artificial intelligence needs to be trained. An AI example is the Google Images model, where the algorithm is trained to identify photos. For instance, for the search engine to find dog photos, the algorithm must be exposed to various images to learn the repeated shapes and patterns of dogs. After analyzing a sufficient number, the algorithm learns how to recognize the ways that define a dog and will be able to identify it in any scenario.

However, this artificial intelligence learning process can have poor results and malicious applications, such as Amazon’s case. The company used hiring algorithms to screen candidates based on the resumes of personnel employed in the last ten years. The problem was that many of those hires were men since it is an industry dominated by that gender. So the artificial intelligence used for this purpose observed this pattern and ended up learning that women “are not good candidates.” Although this case shows how algorithms can increase biases and discrimination, they also identify these same prejudices and bigotry in society. For example, if you search for “student” on Google, your image search results show mostly white people’s faces.

Algorithms in assessments

Education is also not safe from algorithms. With the pandemic, millions of students and teaching staff had to be confined to their homes and continue online classes regardless of whether there was the right infrastructure or relevant knowledge to adapt to the new format.

Among the greatest challenges to pandemic education is evaluation, so many resorted to using algorithms to help with this, as was the case with England. The Office of Qualifications and Examinations Regulation in England (Ofqual) created an algorithm to standardize assessment scoring to prevent grade inflation. This resulted in 39% of students receiving lower grades, setting off mass protests across the country demanding to review the process.

This was not the only problem facing the country as it had a similar problem with the advanced level International Baccalaureate (IB) grades. These assessments are made at the end of the year and define whether students enter college or not. Again, the algorithm lowered many students’ results, causing some students to lose access to higher education. In both cases, the government canceled the results by opting for evaluations based on its mock exams and teacher evaluations. They also allowed students to take their exams later in the year.

The opposite happened in the United States, where several students discovered that the artificial intelligence of the Edgenuity platform had a preference for certain words. Soon, they started using these words in all their answers, and the machine assumed they covered the whole topic and gave them the maximum score.

The problem is that an algorithm that standardizes or assigns grades does not consider many important aspects of a student’s skills. They focus more on whether the students memorized the information well, without recognizing whether they understood or assimilated the content. Still, although there is a long way to go before these algorithms can effectively assess tests, the use of this technology, well-targeted, can be used to benefit education.

How do they use algorithms and artificial intelligence in Japan?

IBM and the Consortium for Renovating Education for the Future (CoREF) seek to transform Japan’s education system. They believe that the cognitive sciences and active and collaborative learning dramatically improve education. With this in mind, they developed the constructive puzzle method of knowledge, dividing students into small groups and getting them to consider a topic from multiple angles. Instead of having a static study plan, the teacher must design lessons focused on each student, observe them, provide feedback about their activity, and collaborate with other teachers to improve.

To enable teachers to monitor their students’ interactions effectively, IBM developed “Watson speech to text” that monitors the quality of students’ conversations, leading to improving the method’s planning and feedback. The teachers teach Watson keywords they expect to emerge during the exercise to evaluate how much they understood. Also, Watson records and transcribes students’ discussions, searching for the teachers’ anticipated keywords to review later and improve the machine’s algorithm.

The purpose of IBM and COREF is to strengthen the students’ capacity for independence and collaboration. They believe that using algorithms and AI makes the teachers focus more on inspiring creativity and the love of learning rather than making the students memorize data to pass a test.

Japan also invested $227,000 in a pilot test to improve English skills in 500 classrooms. The country forces all primary school students and those between the ages of 12 and 15 to learn the language. However, due to the demand, they struggle to find qualified English teachers. That is why the government decided to incorporate programmed robots with algorithms and artificial intelligenc
e to verify each student’s English pronunciation.

In this age where technological advances are increasing rapidly, algorithms have become the basis of many of these innovations. Although there is still a great deal of ignorance in the technical fields about the differences in Artificial Intelligence, Machine Learning, and Deep Learning, algorithms are present. Algorithms are the instruction sets that guide the technology in these areas to fulfill their functions; they help them learn.

This is the reason why there is so much talk about algorithms because they are something that impacts each person’s daily life. It is necessary to educate students to understand what they are and how they work to comprehend where technology is moving and enable them to face the challenges of entering the labor force surrounded by technology.

Translation by Daniel Wetta.

Paulette Delgado

Share

This article from Observatory of the Institute for the Future of Education may be shared under the terms of the license CC BY-NC-SA 4.0