AI BIAS AND DISCRIMINATION

Other aspects of AI that need to be discussed are bias and discrimination, Dr Chowdhury said.

“I think what’s very important for a lot of people… is to understand that this technology is not going to be as ubiquitous as we think, unless we solve the tabletop problems like bias and discrimination ” said the data scientist who runs an organization. provide ethical AI solutions.

She pointed out that AI models are trained from internet data, which is the data of the Western world.

“Thinking about applications in Asia will force us to think about data, sources and biases that might be very Western-focused, or simply inappropriate for use in this region,” she said.

“For example, we are trying to create AI models to improve agricultural techniques. What will happen if these models produce biased results because they do not understand the cultures of the region or the language spoken there?

She added that global priorities for standards and use cases are largely determined by just a few countries.

This means that solutions to improve crop yields due to climate, for example, are not a priority.

Dr Chowdhury highlighted that these are issues faced by farmers in countries that do not have a seat at the same table, such as Bangladesh – where his family is from – and Vietnam.

“It’s really critical that we focus on what’s useful for people and for the rest of us…broadband access, connectivity, mobile access, all of those things are important in an engaged population in the future of AI,” she said.

Funding largely comes from industry, which in turn places it in the position of agenda-setter, Dr Chowdhury noted. She added that philanthropies must play a greater role.