Real AI provides accessibility to all

By Michael Hainsworth

Many of the technologies we take for granted today got their start as assistive technologies for people with disabilities. As far back as the 1800s, the typewriter was invented to help the blind lover of an Italian inventor correspond with him and the world. Today, artificial intelligence-powered speech recognition accomplishes the same task. It’s cited as one the most notable example of technologies that got their start helping individuals with disabilities live better lives.

Microsoft’s Director of Accessibility, Dave Dame, points out that as we age, we all become disabled to one degree or another, “it’s just some of us beat you to it.” In 1971, Dame was diagnosed at birth with Cerebral Palsy. His parents were advised to put him in an institution, believing he’d likely never contribute to society. Fifty years later, he’s at the top of a key division at one of the most powerful companies in the world. “Technology provided the virtual and digital ramps I needed to really give me the career that I have and the opportunities I have,” he says.

Dame adds that features like auto-complete that may seem trivial to most of us help make the world better for someone for whom the most trivial thing is complex. Text prediction has at its heart machine learning, the foundation of artificial intelligence. The more we type, the better AI gets at predicting what we will write next. “For me to type one sentence,” Dame points out, “it’s almost the same effort as typing pages of documents.” Combining assistive technologies takes things to a whole new level. “Technology being able to utilize other senses like voice to text allows me to use what I have better control over to minimize what I don’t have control over and really give me the ability to communicate,” he says.

Artificial intelligence has already revolutionized voice recognition and now it’s taking communication to a whole new level. Experts predict instantaneous AI-powered translation of meetings into other languages is less than 10 years away.

Meantime, AI researchers need to take people like Dame into consideration when building their machine learning algorithms to ensure unintended bias is removed from data sets. He cites image recognition systems as an example: when an AI camera saw him in his wheelchair, the algorithm translated that to a baby in a carriage because that was the closest match for his head and that height. “It really gives us a chance to step back and realize the importance of the uniqueness and differences before we put that in the machine and make us more aware of our own individual biases that we didn’t realize we have.”

But how does Dame describe his view of Real AI for the future?

“Optimistic. Because I believe it’s going to make a better world.”

Related Posts