Computing Bias Lesson 🧠

Introduction 📌

Computing bias occurs when computer systems systematically and unfairly discriminate against certain groups. Bias in computing can reinforce social inequalities, create unfair advantages, and lead to real-world harm. This lesson explores how bias emerges, real-world examples, and strategies for mitigation.

Submission Form Here

What is Computing Bias?

“We can say that a computer system is biased if it both unfairly and systematically discriminates against one group in favor of another.” – Nissenbaum et al. (Cornell Paper)

Doctor or Nurse Drawings 🎨

Types of Bias in Computing 🔍

Pre-existing Social Bias

Technical Bias

Emergent Social Bias

Question 1: ❓

A company implements an AI hiring system that unintentionally prefers male candidates over female candidates because it was trained on past hiring data from a male-dominated industry. What type of bias is this an example of?

Reveal Answer Answer: A) Pre-existing Social Bias This bias originates from societal inequalities reflected in historical hiring data.

Question 2: ❓

An AI-powered chatbot gradually starts using offensive language after interacting with online users who repeatedly expose it to toxic content. What type of bias does this represent?

Reveal Answer Answer: C) Emergent Social Bias The chatbot learns bias over time from user interactions, rather than being biased from the start.

Explicit Data vs. Implicit Data in Computing Bias 🧩

When analyzing how bias appears in computing, it’s important to understand the difference between explicit data and implicit data—both of which can influence biased outcomes.

Explicit Data 📊

Implicit Data 🔍

How Does This Relate to Bias? ⚖️

Bias can emerge when implicit data is used in ways that reinforce stereotypes or favor certain groups unfairly. Since implicit data is often inferred, it can exaggerate trends and introduce bias unintentionally.

For example:

Popcorn Hack #1:

Think of a real-world example where a computer system has shown bias. It could be something you’ve read about, experienced, or imagined.

Question:

Describe the biased system, explain what type of bias it represents (Pre-existing Social Bias, Technical Bias, or Emergent Social Bias), and suggest one way to reduce or fix the bias.

Example Response "A facial recognition system fails to accurately recognize darker-skinned individuals. This is an example of **Technical Bias** because the training data lacked enough diversity. A way to fix this could be to ensure the dataset includes a wide range of skin tones before training the model."*

Real-World Examples of Computing Bias 🌍

Amazon’s AI Hiring Tool

Google Translate Gender Bias

Microsoft’s AI Chatbot Tay

How to Mitigate these Biases?

Pre-existing Social Biases

Technical Bias

Emergent Social Bias

Popcorn Hack #2:

Bias in computing can lead to unfair outcomes, but there are ways to reduce it

Question:

In the financial industry, an AI system used to approve loan applications unintentionally favors male applicants over female applicants because it was trained on past loan approval data, which reflected gender biases. This is an example of Pre-existing Social Bias.

Give two ways to mitigate this bias and make the system more fair.

Example Responses A music app recommends only popular songs and ignores lesser-known artists. - 1️⃣ **Add more variety** to the recommendations. - 2️⃣ **Let users choose** what types of music they want to hear.

MCQ Topic Connection to Computing Bias
2.3: Extracting Information from Data Bias can emerge when datasets are skewed or incomplete, leading to biased AI predictions and decisions.
3.12: Calling Procedures If biased functions or APIs are repeatedly used, the bias spreads throughout a system, such as facial recognition errors.
3.17: Algorithmic Efficiency Some efficient algorithms may prioritize speed over fairness, reinforcing biases in search engines or recommendation systems.
5.1: Beneficial and Harmful Effects Biased algorithms can offer personalized recommendations (benefit) but may also create filter bubbles and reinforce stereotypes (harm).
5.2: Digital Divide Bias in computing can worsen accessibility gaps, such as voice recognition software struggling with non-native accents.

Questions to Ask About Bias (From CollegeBoard)

Enhancing or Intentionally Excluding

Intentionally Harmful/Hateful

Receiving Feedback from a Wide Variety of People

Homework Hack: Understanding Bias in Computing

Question:

Think of a system or tool that you use every day—this could be a website, app, or device. Can you identify any bias that might exist in the way the system works?

Task:

  1. Describe the system you’re thinking of.
  2. Identify the bias in that system and explain why it might be biased. (Is it Pre-existing Social Bias, Technical Bias, or Emergent Social Bias?)
  3. Propose one way to reduce or fix the bias in that system.

Example:

“I use a music app that suggests songs based only on what I already listen to. The system is biased because it limits me to a narrow selection of music. This is an example of Algorithmic Bias. To fix this, I would add an option for the app to recommend songs from different genres that I don’t normally listen to.”

🌟 Extra Credit Opportunity 🌟

Objective:

Complete the game to earn 100% and provide detailed explanations for each bias present in the game. For each bias, explain why it occurs and suggest a solution to eliminate or reduce its impact.

Task:

  1. Play the Game
    Play the provided game and aim to achieve 100% completion. Make sure to pay attention to the reasons that those biases are present.

  2. Identify the Computing Biases:
    As you play, take note of any of the following computing biases:
    • Pre-existing Social Bias: Biases that are inherited from society, such as racial, gender, or cultural biases, and are reflected in the system or game.
    • Technical Bias: Biases that arise due to limitations in the technical aspects of the system, such as algorithmic or data limitations that result in unequal treatment of certain groups.
    • Emergent Social Bias: Biases that develop during interactions between users or players within the system, such as crowd behavior or group dynamics that lead to unfair or unbalanced outcomes.
  3. Provide Explanations and Solutions:
    For each bias you identify, explain why it is a bias and how it impacts our lives. Then, suggest at least one solution to address or minimize that bias. Solutions could include:
    • Altering code
    • Introducing more diverse backgrounds
    • Modifying the narrative or decision-making process

Example:

Extra Learning Links: Computing Bias

🌐 Websites

🎥 Videos

📊 Academic References & Real-World Examples

📄 Definition and Information

🔍 Real-World Examples

🖥️ Interactive Resources