Smoke that codes: Women in tech, AI, & evolution of power | Krystal Trashawn Morris | TEDxChinotimba

Quick Overview

The speaker argues that the current development of Artificial Intelligence (AI) and robotics, driven largely by data and decisions made without sufficient representation from women, leads to biased outcomes, and proposes a five-step framework to ensure diversity and representation are integrated into technology development for a more equitable future.

Key Points: The speaker emphasizes that AI and robotics learn from human data and decisions, which currently exclude half of humanity (women), leading to systemic bias. Pioneers like Ada Lovelace (who wrote the first algorithm in the 1800s) and Grace Hopper (who invented the compiler) laid essential groundwork for technology that was often not fully recognized. The speaker notes that technologies like social media and even transportation and irrigation are shaped by this limited perspective, resulting in unfair outcomes. The proposed solution is a five-step framework designed to integrate diversity and representation into the creation and teaching of AI/robotics. Step 1: Normalize AI and robotics as early as possible, making it a core subject in classrooms. Step 2: Educate the teachers on technology so they can effectively guide students. Step 5 emphasizes that diversity and representation must be a quality control measure, not an afterthought, to ensure AI reflects the fullness of humanity.

Context: Krystal Trashawn Morris presents at TEDxChinotimba on the critical need to address the biases embedded within Artificial Intelligence (AI) and robotics systems. She establishes that because these technologies learn from existing human data and decisions, the historical exclusion of women—and specifically Black women—results in biased outputs that fail to serve all of humanity, citing historical figures like Ada Lovelace and Grace Hopper as examples of vital, yet often overlooked, female contributions to computing.

Detailed Analysis

The speaker passionately argues that the current trajectory of AI development is flawed because it is trained on data and decisions that disproportionately reflect the perspectives of only half of humanity, specifically excluding women. She uses historical examples, such as Ada Lovelace writing the world's first algorithm before computers existed and Grace Hopper inventing the compiler while asking why computers couldn't understand human language, to illustrate that vital contributions from women have historically been marginalized or overshadowed. She points out that this lack of inclusive data leads to biased technology in everyday life, from social media algorithms to potentially dangerous outcomes in fields like medicine and transportation. To correct this, she outlines a five-step framework: 1. Normalize AI and robotics education early. 2. Educate the teachers first. 3. Teach students how to program robots to control devices. 4. Teach computational thinking (cause and effect, sequencing, visualization). 5. Ensure diversity and representation are mandatory quality control measures, not afterthoughts, so that the future vision of AI reflects the full spectrum of humanity. The core message is that if the creators and the data are not diverse, the resulting technology will inherit and perpetuate those biases, making inclusive education and design imperative for equitable technological evolution.