Coded Bias

Chapter 1: Introduction

  • Excitement to Meet Humans

    • The speaker expresses enthusiasm for meeting humans and learning from their experiences.

  • Background in Computer Science

    • The speaker pursued computer science for its detachment from real-world problems.

    • Attended MIT to create cool technology, focusing on art projects involving computer vision.

  • Aspire Mirror Project

    • Developed a project called the Aspire Mirror to inspire positive self-image by reflection.

    • Utilized computer vision technology to track faces and overlay images, but faced challenges with detection.

    • Discovered bias due to lack of representation in training datasets, leading to the realization about bias in technology.

  • AI and Pop Culture Connections

    • Discusses how AI is often portrayed in Hollywood (e.g., Terminator, Star Wars).

    • Clarifies that current AI is "narrow AI" focused on specific tasks, not general intelligence.

  • Dartmouth Conference and Chess

    • The launch of AI as a formal field initiated by a meeting at Dartmouth in 1956 where chess ability was a benchmark for intelligence.

    • Example of Garry Kasparov vs. IBM's Deep Blue highlights rigid definitions of intelligence.

  • The Impact of Bias in Technology

    • Personal experiences reflect larger systemic issues in AI related to gender, race, and bias in algorithms.

    • Various facial recognition systems showed significant bias against women and darker skin tones.

  • The Role of Data

    • Data is destiny: datasets reflecting historical biases can lead to discriminatory outcomes in AI applications.

    • Importance of monitoring AI for bias to prevent unintended discrimination in society.

  • Algorithmic Powers and Accountability

    • AI's impact on people’s lives raises concerns about accountability and fairness.

    • Emphasis on the asymmetrical power dynamics in technology deployment.

  • Social Awareness and Activism

    • Kathy O'Neil's discussion on AI's societal dangers motivated the speaker to explore activism in algorithmic bias.

  • Conclusions

    • The urgency for awareness and avenues for redress in the governance of emerging technologies.