A Note on Biases in the Era of Intelligent Machines

Published on 17 June 2020

What is bias?

Bias is defined as “a tendency, inclination or prejudice against or towards an individual, a group or a belief.”  It is, in fact, human to have a strong preference for or against something.

Are biases positive or negative?

Biases can be positive or negative. Choosing to eat healthy foods and avoiding unhealthy foods is a positive bias. Negative bias based on stereotypes like race, gender, color, ethnicity, religion, or sexuality of individuals rather than actual knowledge of individuals or circumstances will lead to misunderstanding, rash decisions, social inequality, discriminatory practices, and open hostility. It can have harmful real-world outcomes. Paying attention to positive biases, while keeping the negative biases in check, requires a delicate balance, self-protection, and empathy for others.

Who should learn about biases?

Everyone. Why should biases only be studied by psychologists and behavioral economists? It is good to be aware of biases from an early age. In the 21st century, anti-bias training and being culturally aware are crucial for student survival.

What is cognitive bias?

The most common type of bias is a cognitive bias. Repeated patterns of thinking that lead to inaccurate or unreasonable conclusions is called cognitive bias. 

Why is cognitive bias important?

Learning about cognitive bias is a valuable skill for the 21st century. It is not enough just to know that this bias exists in the world. If we want to be able to fix it, we need to understand the mechanics of how it arises in the first place. 

Artificial intelligence and machine learning are leading the automation era. Soon many jobs will be automated. To solve a problem, data is first collected. Our ethics and biases can influence how data is collected, prepared, and analyzed. If biases creep in, it will not be a true representation of reality. It must not reflect our prejudices. To avoid the predatory behavior of algorithms, which may not understand true intentions, fairness in machine learning is important. Computer scientists create a deep-learning model to solve a problem. If they are biased and unfair, their attitude might creep into their computing and coding.

What are some common cognitive biases?

Everyone should learn about cognitive biases, especially researchers, teachers, and students. For those interested, there is a visual representation of all cognitive biases called Cognitive Bias Codex online. Some common cognitive biases are explained below.

Due to growing up in different places, people have varied perspectives and preferences. Due to lifelong societal inputs, people may behave in a prejudiced manner towards others, unknowingly, automatically, and due to their conditioned brains making quick judgments and assessments of people and situations. This is called unconscious bias or implicit bias. Some people, to avoid being stereotyped, will go to great lengths to avoid conforming to a particular group. This is called stereotype threat.

Confirmation bias is our tendency to search for and focus on the information that supports what we already believe while ignoring relevant facts that go against those beliefs. Groupthink happens when people in a group tend to agree with each other to maintain harmony, without critical evaluation of decisions. Anchoring bias is when a group of people just agree on the first piece of information they get to make decisions.

Response bias is a tendency for people to respond falsely to questions. Information bias is when key information is interpreted inaccurately. Attribution bias is when you attribute reasons or motivations to the action of others without concrete evidence to support such assumptions. Commercial bias or political bias is when a journalist is influenced by a wealthy owner of a newspaper or shareholders, or a television news channel is influenced by the political party which owns it.

In scientific research, unfair sampling of a population or inaccurate estimation of average will lead to statistical bias. A researcher will have to be aware of selection bias to prevent errors that might occur when selecting who or what is going to be studied. People tend to look for patterns in random data. This is known as the clustering illusion. When the statistical significance of results influences how research is reported, it is called reporting bias. When a researcher influences the participants in a study, it is called the observer effect.

People tend to believe others, until evidence to the contrary surfaces. This is called truth bias, which enables society and commerce to run efficiently. Truth bias will make people believe what they hear, see, or read. So much so, they will turn a blind eye to the loose ends that do not fit into the narrative. Even when some information is withheld in a story, the story remains true.

In these times of fake news, misinformation, and disinformation, one has to be alert to what we consume online. Do not take things at face value. Read books and journals, watch documentaries, and listen to different sources of news from around the world. Listen to people when they talk. Are they conveying facts or are they trying to convince you of the facts? Be a little skeptical. Challenge assumptions. Be open. Talk less. Listen more.

Are you aware of the 17 Sustainable Development Goals (SDGs)?

Look at the image below to know what SDGs are all about.

Source: United Nations Sustainable Development Goals
The content of this publication has not been approved by the United Nations and does not reflect the views of the United Nations or its officials or Member States.

What is the connection between SDGs and biases?

Being aware of our biases, especially cognitive biases, will help us understand our behaviors. By making suitable behavioral changes, we can work towards SDGs set by the United Nations in 2015 and achieve it by 2030 on a global scale. It begins by getting to know ourselves. So, get to know yourself.

Excerpted from the book – The Digital Pa-rant (2020) – by Sivani Saravanamuttu

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s