The national unemployment rate for those with disabilities is 70%, even though it’s been 30 years since the passage of the Americans with Disabilities Act. But there remain huge opportunities for this community to participate more fully in the world of work. While government-sanctioned and -sponsored efforts have addressed disability employment challenges for a long time, now the country needs business leaders to better leverage this reservoir of untapped talent. They can not only make companies more inclusive and fair, but also help drive digital transformation efforts. So the CDX digital transformation community recently brought together a group of disability empowerment leaders to map out where we stand and where we might go.
The conversation featured Guy Tonye, chief engineer at cross-platform voice app developer and integrator Zammo, John Robinson, CEO of Our Ability Inc., a unique company that specializes in helping disabled people find employment, and Bill Curtis-Davidson, AI Lead for the Partnership on Employment & Accessible Technology (P.E.A.T.) initiative at the U.S. Department of Labor. These three experts explored how voice technologies and in particular artificial intelligence (AI) provide unique new opportunities for workers with disabilities. All three are strong advocates of increasing access and opportunity for people with disabilities and ensuring that emerging technologies are accessible to all workers. And they all believe it’s possible to make real progress, but that most businesses are just not trying hard enough.
The Department of Labor calculates that while 67% of non-disabled people participate in the workforce, only about 21% of people with disabilities do. And of those who want to work, the unemployment rate for disabled people is over 11%, compared to 6.9% for the general population as of October (and that is a figure far higher than usual because of the pandemic).
Our Ability serves as an employment conduit for people with disabilities, and aggressively employs AI to match people with jobs. Robinson, who founded the company, is himself disabled. He has won numerous awards for his entrepreneurship and activism in and for the disabled community. He says the company’s primary mission is lowering the unemployment rate for people with disabilities through the use of technology. Our Ability’s website offers an AI-powered chatbot specifically designed to take disabled people through an extensive dialogue to determine what kind of work they would want, what they may be qualified for, or what kind of training they could benefit from. Explained Robinson: “We know that technology is a great equalizer for people with disabilities and with voice and conversational AI, we can do a better job in harnessing their ability and skills – early and faster.”
AI also increasingly plays a role on the job itself, the panelists said, helping disabled workers by modifying computer usage so they can better see, hear, and reason. AI and other software can make it easier to interact with devices, for example, by enabling the rapid and accurate transcription of speech, and by predicting what text a disabled writer may input.
Curtis-Davidson from P.E.A.T., a public-private partnership funded by the Department of Labor’s Office of Disability Employment Policy (ODEP), says that the pandemic crisis has further helped technology adapt to the needs of the disabled. “This COVID period has done a great job in honing our craft with AI,” he said, “in training and bringing people with disabilities together with businesses.” P.E.A.T’s recent annual Think Tank event was all about the virtual workplace. Curtis-Davidson called it “a great convening of multiple stakeholders from academicians to researchers, technology leaders, disability-led organizations, and advocacy groups.” Among its other roles, P.E.A.T. helps employers think about how to install and use accessible telework technology.
Advocacy is needed continually, the panelists said, not only to further inclusive efforts in the workplace but to counteract the deployment of technology across society that discriminates or fails to include the disabled. In March of 2019, the AI Now Institute at New York University, with the NYU Center for Disability Studies and Microsoft, issued a report that documented AI disability bias. “Take, for example, the use of AI in developing autonomous vehicles,” the report explained. “If the data used to train a pedestrian recognition system doesn’t include representations of people using scooters or wheelchairs, it’s likely that such people won’t be “recognized” as pedestrians.” The report followed a major conference that gathered disability scholars, AI developers, and computer science and human-computer interaction researchers to discuss the intersection of disability, bias, and AI.
How society responds to impairment can often compound the challenges created by physical or mental disabilities themselves. P.E.A.T.’s Curtis-Davidson pointed out that care must be taken with the source and nature of the data used to train the AI: “We have to be sure that our algorithms used in AI and machine learning account for diversity and inclusive data sets.” He says concerns for disability “should be included in our collective efforts to reduce algorithmic bias reduction, which is also at work in other areas of diversity, such as race and gender.”
Zammo’s chatbots use both voice and text technology, specifically to enable accessibility. It has a variety of application development tools meant for designing apps for disabled users. Said Tonye: “My focus is on how can we bridge the gap between the technology and the end user.”
Tonye is adamant that the enabling technologies are available to make a fundamental impact on driving disability inclusion now. He and the other speakers noted that Microsoft in particular has worked hard to ensure its software is based upon what Curtis-Davidson called “inclusive design.” Tonye is optimistic: “It’s simply changing how people access and convey information, and our job is to help drive that change. We can solve these problems.”