Problem 1: Shallow Data Pool
Text-to-image AI models need huge amounts of image data and image annotation to turn your text into an image. Specifically, it requires captioned pictures so the AI can learn how to process your request. It’s like asking someone to draw something in Pictionary. They recall what it looks like from experience and reproduce it on paper. The better the drawing, the easier it is to correctly identify.
The problem is huge quantities of text-to-image AI data are coming from the web. As we know there are some areas of the web which contain inappropriate and unsavoury material. Datasets tend to reflect social stereotypes, oppressive viewpoints, and derogatory, or otherwise harmful, associations to marginalized identity groups.
Better filtration can remove such content.
Problem 2: Outdated Outputs
There may be replication of prevailing social biases and stereotypes. For example, an overall bias towards generating images of people with lighter skin tones and a tendency for images portraying different professions to align with Western gender stereotypes.
The output, therefore, is often racist, sexist, or toxic in some way. In early releases of the AI models, if you ask it to generate images of a “flight attendant”, almost all the subjects will be women. Ask for pictures of a “CEO” or even a “lawyer” and you see white men.
There are 7 ways to minimise bias:
- Be aware of contexts in which AI can help correct for bias and those where there's a high risk for AI to exacerbate bias.
- Look for hidden stratification, and react appropriately.
- Establish processes and practices to test for and mitigate bias in AI.
- Engage in fact-based conversations about potential biases in human decisions.
- Fully explore how humans and machines can best work together.
- Invest more in bias research, make more data available for research while ensuring privacy, and adopt a multidisciplinary approach.
- Invest more in diversifying the people operating in the AI industry.
Collaborate with us