Cultivating more intelligent government has long been a hallmark of public sector reform, as more agile and learning-based forms of governance are essential to innovation. The rise of Artificial Intelligence (A.I.) is an extension of this logic – albeit with one important difference. In place of human ingenuity, the machines take charge. Well, perhaps not take charge (not yet, at least…), but A.I. essentially implies intelligence exhibited by machines.
In a world evermore saturated by data, A.I. is more a necessity than a luxury. Asking a human to extract value-adding patterns from today’s dataflows without technology would be akin to dispatching a lone diver in an ocean to understand the migratory patterns of fish.
Accordingly, technology companies are investing heavily in A.I. While Facebook has hired thousands of employees to filter out fake news, such a gargantuan task invariably requires automation. Healthcare is being revolutionized not only through the automation of routine procedures, but also the dissection of rapidly expanding patient data pools (the UK now asks citizens to opt out of such sharing rather than opt-in).
For core government, proponents of A.I. point to routine automation and customer service innovation as hallmarks of a new cognitive state. A recent report by Deloitte Consulting, for instance, envisions five potential benefit streams: i) overcoming resource constraints; ii) dramatically cutting paperwork; ii) reducing backlogs; iv) improving prediction; and v) answering citizen queries.
Experimentation is growing. A Singaporean initiative entitled, ‘Conversation as a Platform’ seeks to re-imagine interactions between government and citizen through three phases: the use of chatbots for simple, factual questions; assisting citizens with the completion of routine transactions through self-service and online channels; and eventually, service customization to enhance the user experience and improve outcomes.
Following the lead of several banks in that country, Australia’s federal government has also experimented with service improvement through forms of A.I. such as chatbots and predictive analytics. Unfortunately, ensuing controversies stemming from some early initiatives point to the sorts of pitfalls that wait.
The ‘robo-debt’ scandal, for example, stems from efforts to reduce fraud and collect outstanding debts through automation. Many citizens were incorrectly notified of debts they had not incurred, while others received pro-active warning letters hardly suggestive of bettering citizen-centric service. In June, a Senate Inquiry called for the program’s suspension until problems could be fixed: the Australian Council of Social Service has called for outright abolishment.
Less dramatically, the Australian Tax Office has deployed a chatbot named ‘Alex’ that has successfully replied to more than one million inquiries by citizens. Many State governments in the US have also begun to trial chatbots for information inquiries, in attempt to lessen and better target the need for direct human interaction.
Such automation invariably gives rise to concerns about job losses – while some groups champion human worker displacement as a worthy cause. One recent report by a British think tank, Reform, suggests that over the next fifteen years up to a quarter of a million public servants could cede their functions to robots of one sort or another. The result, according to Reform, could be a more efficient and nimble public service.
Longer term, there are also growing questions about a future where A.I. eclipses the ability of humans to keep up. A notable critic in this regard is Tesla founder Elon Musk who told students at MIT: “I think we should be very careful about artificial intelligence…I’m increasingly inclined to think that there should be some regulatory oversight, maybe at the national and international level, just to make sure that we don’t do something very foolish.”
In Canada, the recent federal budget featured new investments into A.I. research and commercialization, while many proponents of this burgeoning field view President Trump’s anti-immigration rhetoric as a further opportunity for this country to entice human capital. The Senate, moreover, has embarked upon a study to better understand A.I. and robotics in health care.
As much as such efforts matter, what’s missing is a corresponding emphasis on modernizing core government operations – in a way that seeks to balance both opportunity and risk. In an era where the federal government remains hamstrung by its faulty payroll and email systems, a lagging public sector constraints our collective intelligence, artificial or not.
Jeffrey Roy is a professor in the School of Public Administration at Dalhousie University (roy@dal.ca).
