Welcome to ADB - a data base against Data Bias! We are creating awareness of data bias in AI and are building a platform to provide free access to unbiased AI data training sets, accessible to anyone who is building technology employing Artificial Intelligence.
Albeit innovative technologies employing AI have the potential to save humanity with intricate solutions they also pose risks. Data Bias in AI training data is one of them and we want to mitigate that.
Our world is crisscrossed with a multitude of AI applications invisibly woven into the fabric of our daily life. Sophisticated data-based technology with decision-making power based on AI is embedded in delicate areas such as education, hiring, health care, and criminal justice and effects individuals and organizations across all sectors while touching all areas of life.
Highly biased outcomes, or anomalies in the output of machine learning algorithms, question the validity and justice of AI-powered solutions and have brought a tidal wave of complex issues that will not abate.
Bias as a concept is defined as the ”systematic discrimination against certain individuals or groups of individuals based on the inappropriate use of certain traits of characteristics.”
Elisabeth L'Orange, Co-founder of Oxolo, and Sydney Otten Co-founder of Straiqr, both deep tech companies focussed on AI, set up ADB together with scientists specialized in AI, investors and simply humans who care about equality. Many of us have experienced Data Bias firsthand. We want to change it by providing a solution:
Creating a platform to share unbiased data sets and ultimately producing our own unbiased training data sets.
Let us know if you want to hear from us!