How to keep human bias out of AI | Kriti Sharma

How to keep human bias out of AI | Kriti Sharma

How AI Decisions Impact Our Lives

The Role of AI in Decision Making

  • The speaker, an AI developer, highlights the prevalence of AI in making decisions about individuals' lives, prompting reflection on how many such decisions have been made recently.
  • Media often sensationalizes AI's potential threats, but the speaker emphasizes that the real concern lies in biases embedded within AI systems rather than fictional narratives.
  • Examples are given of biased assumptions made by AI regarding race and gender, illustrating how these biases can lead to discriminatory outcomes in various sectors.

Real-World Implications of Bias in AI

  • The speaker presents alarming examples of biased decision-making by AI: assumptions about loan repayment based on race and job suitability based on gender.
  • These biases affect critical life aspects such as job interviews, insurance rates, credit scores, and performance reviews—often without transparency or accountability.

Gender Discrimination and AI

  • A scenario is described where an AI trained predominantly on male candidates perpetuates gender bias by favoring male applicants for tech roles.
  • The speaker argues that while human discrimination is unacceptable, similar actions by machines are often overlooked due to their non-human nature.

Societal Influence on Perception of Gender Roles

  • Voice assistants like Siri and Alexa are discussed as embodying traditional gender roles; they are typically designed as female figures serving users' needs.
  • This design choice reinforces outdated stereotypes about women being subservient while high-powered AIs tend to be male-oriented.

Solutions for Ethical AI Development

  • Despite the challenges posed by current biases in technology, there is hope. The speaker outlines three actionable steps to improve ethical standards in AI:
  • Be aware of personal biases and those present in technology.
  • Ensure diverse teams contribute to technology development.
  • Provide diverse experiences for AIs to learn from.

Personal Experiences with Bias

  • The speaker shares a personal anecdote about facing skepticism as a woman in tech forums when using her real identity versus a neutral profile picture and name.
  • This experience underscores systemic elitism within tech fields that favors certain demographics over others regardless of actual skill or knowledge.

Importance of Diversity in Tech

  • By hiding her identity online, the speaker was able to engage more effectively without facing patronizing comments related to her gender.

Creating Inclusive AI: The Role of Diverse Perspectives

The Need for Storytellers and Problem Solvers

  • Emphasizes the importance of diverse backgrounds in developing AI personalities, highlighting the need for storytellers and problem solvers to address real issues.
  • Advocates for building technology that focuses on positive outcomes rather than fears of job loss or racism in AI.

Real-World Applications of AI

  • Discusses potential life-changing applications of AI, such as providing remote medical diagnoses to pregnant women in rural areas.
  • Highlights how AI can assist victims of domestic violence by offering discreet access to legal and financial advice.

Optimism About Technology's Future

  • Expresses a hopeful outlook on technology's ability to create a more equitable world despite prevalent narratives about existential risks from robots.
  • Stresses the necessity of inclusive development practices, advocating for representation across genders, races, and backgrounds in tech creation.

Addressing Bias in Technology

  • Urges awareness regarding biases embedded within technology, using examples like targeted ads that reflect societal stereotypes.
  • Encourages individuals from all backgrounds to engage with AI development, emphasizing that expertise is not limited to traditional profiles like Mark Zuckerberg.

Call to Action for Education and Inclusion

Channel: TED
Video description

AI algorithms make important decisions about you all the time -- like how much you should pay for car insurance or whether or not you get that job interview. But what happens when these machines are built with human bias coded into their systems? Technologist Kriti Sharma explores how the lack of diversity in tech is creeping into our AI, offering three ways we can start making more ethical algorithms. Get TED Talks recommended just for you! Learn more at https://www.ted.com/signup. The TED Talks channel features the best talks and performances from the TED Conference, where the world's leading thinkers and doers give the talk of their lives in 18 minutes (or less). Look for talks on Technology, Entertainment and Design -- plus science, business, global issues, the arts and more. Follow TED on Twitter: http://www.twitter.com/TEDTalks Like TED on Facebook: https://www.facebook.com/TED Subscribe to our channel: https://www.youtube.com/TED