Smarter together: Why artificial intelligence needs human-centered design

Seekers after the glitter of intelligence are misguided in trying to cast it in the base metal of computing. Artificial intelligence (AI) has emerged as a signature issue of our time, set to reshape business and society. The excitement is warranted, but so are concerns. At a business level, large "big data" and AI projects often fail to deliver. Many of the culprits are familiar and persistent: forcing technological square pegs into strategic round holes, overestimating the sufficiency of available data or underestimating the difficulty of wrangling it into usable shape, taking insufficient steps to ensure that algorithmic outputs result in the desired business outcomes. At a societal level, headlines are dominated by the issue of technological unemployment. Yet it is becoming increasingly clear that AI algorithms embedded in ubiquitous digital technology can encode societal biases, spread conspiracies and promulgate fake news, amplify echo chambers of public opinion, hijack our attention, and even impair our mental well-being.2 Effectively addressing such issues requires a realistic conception of AI, which is too often hyped as emerging "artificial minds" on an exponential path to generally out-thinking humans.3 In reality, today's AI applications result from the same classes of algorithms that have been under development for decades, but implemented on considerably more powerful computers and trained on larger data sets.