Artificial Unintelligence

In the current debate about artificial intelligence (AI), ‘big tech’, and data, Meredith Broussard stands out as a remarkably clear and eloquent voice. In Artificial Unintelligence (2018), she aims to empower people by explaining what computers and algorithms do, who designs them, and who stands to benefit from the rise of technology.

Maths rather than intelligence Broussard’s recurring message is that technologies – both hardware and software – are created by humans. Simple as this observation may seem, in practice, there is a great deal of misunderstanding about what we can realistically expect from digital technology. The misconceptions are fuelled by promises from the industry. Our language for computers is also misleading. Since computers neither know nor think as sentient beings do, ‘intelligence’ is an inaccurate term. Instead, they consist of multiple layers operating on mathematical principles. The same applies to machine ‘learning’, which essentially means that a machine can improve at its programmed tasks, not that it acquires knowledge, wisdom, or agency. This is not genuine ‘learning’.

Statistics on steroids When we talk about AI, we mean narrow AI – ‘statistics on steroids’, as Broussard calls it. Narrow AI is designed to perform specific tasks or a set of tasks. General AI, on the other hand, exists only in Hollywood movies. The industry continues, however, to dangle the illusion of a future that was never possible to begin with.

Technochauvinism is often accompanied by fellow-traveler beliefs such as Ayn Randian meritocracy; technolibertarian political values; celebrating free speech to the extent of denying that online harassment is a problem; the notion that computers are more “objective” or “unbiased” because they distill questions and answers down to mathematical evaluation; and an unwavering faith that if the world just used more computers, and used them properly, social problems would disappear and we’d create a digitally enabled utopia.

Human-in-the-loop systems Broussard draws on her coding background to illustrate the risks of treating data – again, generated by people – as immutable truth. While AI can detect patterns too complex for humans, it cannot account for everything. This makes fully automated decision-making undesirable; instead, the author argues that human-in-the-loop systems, which combine AI with human intelligence, are a more realistic path forward. She also highlights that most engineers today are not creating new technologies, but maintaining existing ones.

Big tech Towards the end, Broussard shifts focus to the people behind the big tech industry, criticising their homogeneous backgrounds and indifference to social issues.

we have a small, elite group of men who tend to overestimate their mathematical abilities, who have systematically excluded women and people of color in favor of machines for centuries, who tend to want to make science fiction real, who have little regard for social convention, who don’t believe that social norms or rules apply to them, who have unused piles of government money sitting around, and who have adopted the ideological rhetoric of far-right libertarian anarcho-capitalists. What could possibly go wrong?

Artificial Unintelligence is often cited, and I understand why: Broussard conveys her expertise in an accessible way and takes a stance. Some sections are so detailed that I skimmed through them, but in the end, I learned a lot. I will continue my reading on digital technologies by revisiting Weapons of Math Destruction by Cathy O’Neil.

We need to stop fetishizing tech. We need to audit algorithms, watch out for inequality, and reduce bias in computational systems, as well as in the tech industry.