Firms Target Autistic People for AI Jobs
Firms scouting out artificial intelligence development talent are targeting a very peculiar subset of the population: those with autism. Large companies such as Credit Suisse, Dell, Microsoft, DXC and Ernst & Young have all established neurodiversity programs for this very purpose, with a dozen Fortune 500 companies being advised to do the same.
But why target autistic people, you may ask? The answer comes from the fact that autistic employees are usually hyper-focused and very analytical thinkers, with a superb grasp of technology. They find it easier to work long, gruelling hours on repetitive artificial intelligence tasks, such as labelling images for computer-vision systems, without getting bored. Furthermore, they are naturally good at pattern recognition, allowing them to systematically develop and try out artificial intelligence models with relative ease.
Unsurprisingly, demand for workers skilled in artificial intelligence, data analysis and related areas has exploded in recent years, with the IT jobless rate being just 1.3% in May 2019 (CompTIA), and as a result ramping up competition for scarce labour. On the other hand, a large quantity of autistic adults are unemployed, with around 42% of autistic people who were in special education in high school having no paid jobs in the initial six years after graduating high school (Drexel University 2015).
Scheiner says firms usually find autistic employees through such organisations as her’s, but some can be socially awkward and introverted, which means that firms have to make a few accommodations. For instance, EY and DXC don’t conduct interviews with them due to the fact that those with autism usually do poorly in unscripted social situations, instead implementing performance assessments over several weeks.
Currently EY has around 80 autistic workers, more than twice the amount they had in 2018, who work across the firm’s five USA neurodiversity centres of excellence. Last year in Dallas, a 14-person team (eight of which are on the autism spectrum) created an algorithm to automate the making of EY consulting contracts. It ended up producing 2000 contracts per month and saving the company about 500,000 work hours per year. The team also made neural networks to rapidly locate potential areas for tax deductions for a client, taking only 12 minutes to process five years worth of memos, emails and other miscellaneous paperwork.
Nine months ago, Credit Suisse commenced a neurodiversity hiring program at its Research Triangle Park office in North Carolina. As of writing, it has two autistic employees and is about to hire a third, and also has seven going through a 12-week apprenticeship program. The two currently in action are creating machine-leaning algorithms to upgrade the client service for the company’s investment banking business.
Kenneth Johnson, one of the aforementioned employees, is a robotics engineer in global markets technology process automation and quality control. One element of the job is applying machine learning models to extract and classify data from unstructured emails and attachments to improve a database that is used by the bank’s traders. Mr Johnson claimed he brings “a new perspective to my team” and that he works with “exceptional speed and efficiency”.