From relentless automation to algorithmic bias and human rights abuses, artificial intelligence (AI) has a laundry list of well-known potential costs and risks that do not bode well for the future of inclusion.
Take Amazon’s recent foray into mass surveillance, for example. Designed by Amazon Web Services, Amazon Rekognition is a deep learning AI that makes it easier for businesses to add image and video analysis to their applications. Common use cases include searchable image and video libraries, face-based user verification, and unsafe content detection. But documents obtained by the American Civil Liberties Union (ACLU) of Northern California show that Amazon is now supplying its facial recognition technology to governments and law enforcement agencies, as well.
Last month, a coalition of over 70 civil liberties and human rights organizations signed a letter that calls on Amazon to stop powering the government’s surveillance infrastructure. The coalition argues that Rekognition’s real-time tracking and monitoring features are “a powerful surveillance system readily available to violate rights and target communities of color.” In the hands of governments, the technology is “primed for abuse” and a “grave threat” to civil rights and liberties.
This raises an important question: is it possible to foster inclusive communities when the government can track the location and behaviour of anyone walking down a street? Unfortunately, Amazon Rekognition is just one example. There is now a growing list of examples that collectively raise a more fundamental question: will advances in AI help include the excluded or further reinforce the barriers of exclusion?
Our current set of policies and institutions will almost certainly result in the latter. Scant regulation and oversight, in addition to a workforce unequipped with the skills for the jobs of tomorrow, will result in greater inequality, discrimination, and exclusion.
To achieve greater inclusion and maximize the social impact of AI, we need innovative and forward-thinking public policies — now.
A good place to start is to define the nebulous idea that is inclusion.
A year ago I had the opportunity to work on a report that examined the language of inclusion and what it takes to build an inclusive society. Written for the 2017 6 Degrees Citizen Space, the report, All of Us: What We Mean When We Talk About Inclusion, was a seminal piece that was indicative of the zeitgeist of the time — for what better defines 2017 than the debate over inclusion? From the Women’s March on Washington to the Charlottesville white supremacy rally, the debate over whether and how we should be inclusive dominated all sides of the political debate in 2017.
I recently reread the report and found it to be just as informative and germane as the week it came out. Sarmishta Subramanian, editor-in-chief of the Literary Review of Canada and author of the report, eloquently captured how inclusion is an ideal that has consistently challenged, inspired, and eluded us — regardless of time or locality. From Ancient Rome to modern-day India, from corporate workplaces to Harvard’s admissions team, she explored a number of examples to unpack what exactly we mean by “inclusion.”
In the closing paragraphs of the report, she provides an answer. She concludes that inclusion is:
Fundamentally about how we see our place in the world, about our ability to imagine and achieve a good life in every area that is meaningful to us. The deprivation or confinement of this ability limits the richness of that life; its expansion sustains that vision, allows it to flourish. The capacity of all citizens to have this, in turn, allows a society to flourish.
The essence of Sarmishta’s definition is that inclusion is our ability to participate in the areas of life that are meaningful to us — be it political, economic, cultural, or social. The obstruction of this ability is what drives exclusion and what causes negative repercussions for individuals and societies.
From this perspective, what inclusion will be in the age of AI is the same as it is today and the same as it was in the past: the ability to imagine and achieve a good life in every area that is meaningful. However, how inclusive the age of AI will be is still up to debate. And, this is the key point, how inclusive the future will be does not depend on the state of AI’s sophistication, but on the policies, regulations, and norms that govern the use and development of AI.
AI Policy for Inclusion
In the past year, over 15 countries have released national strategies to promote the use and development of AI. They almost all include multi-million dollar investments in basic and applied AI research, initiatives to encourage the uptake of AI across the economy, and steps to develop and attract AI talent. Many countries also seek to become the “global leader” in specific areas of AI: the EU wants to set the global standards for AI ethics, China wants to be the world’s primary AI innovation center, and Canada wants to be the global leader in AI research and training.
India, the latest country to release a national strategy, largely follows this mold. Developed by NITI Aayong, the government’s policy think tank, the report recommends creating new research centres for AI, funding national AI fellowships, investing in strategic sectors, and setting up councils to define ethical standards. Unlike other national strategies, however, each of India’s initiatives are geared towards ensuring social and inclusive growth. As Amitabh Kant, CEO of NITI Aayong, argues, “India, given its strengths and characteristics, has the potential to position itself among the leaders on the global AI map — with a unique brand of #AIforAll.”
India is not the first country to connect social progress and inclusion to AI. Canada and France, for example, recently announced ahead of the 2018 G7 Summit a new task force to develop an international study group on inclusive and ethical AI. Japan’s strategy, likewise, focuses on the industrialization of AI solutions for social problems that Japan and the world faces. But India’s strategy does this and more: from research to open data and skills development, every initiative centers on how AI can be used for social development and inclusive growth.
India’s strategy has not been implemented yet, so it is too early to identify best practices for other countries. However, countries should take note of India’s goal of “inclusive technology leadership.” AI can be used to increase productivity, competitiveness, and economic development, but it must also be used to enhance the ability of every person to actively and fully participate in all aspects of life that are meaningful to them.
From Indigenous rights to gender equality, from cleaner water to energy conservation, AI technologies have a lesser-known list of potential benefits and opportunities for the future of inclusion. It is the responsibility of governments to invest in these benefits to ensure that the age of AI is inclusive of everyone.
— — — — — — — — — — — — — — —
Tim Dutton is an AI policy researcher based in Canada. He is the founder and editor-in-chief of Politics + AI. He writes and edits articles for Politics + AI’s Medium page and provides contract work to governments and companies looking to learn about the emerging political risks and opportunities of AI. You can follow him on Twitter and connect with him on LinkedIn.