Code Dependent: Will AI eclipse human agency?

Much has been written about and much will get written about Artificial Intelligence (AI) that is fast revolutionising the way we work, study, and consume. In this rapidly changing world of automated decision-making that is inexorably influencing our behaviour, what it means to be human remains an unanswered question. Delighting us with its humanlike ability to write, visualise and talk back, AI has crunched large data to free us from our own actions. Come to think of it, the machine is taking back our agency and, in some cases, our self-respect. Proponents of AI may loathe the idea of protesting technology’s dramatic progress and present the argument that it is early days still.

However, the glaring fact is that a limiting of the human ability to decide and direct its actions is a fundamental contradiction to our evolution as a species. Little is known about how machine-learning models draw statistical inferences from volumes of data at their disposal. With quantum computing set to exponentially enhance the machine’s ability to intuitively mimic processes, the advent of AI raises more questions than answers.

An absorbing and engaging narrative, Code Dependent is not about how technological infusion is empowering algorithms but is a treatise on how AI has insidiously entered our lives and altered the very experience of being human. With over a decade of experience in writing and reporting on AI for the Financial Times in London, Madhumita Murgia explores what the rise of AI means for us as a society. It is the story of AI but from a human point of view, told through the lived realities of nine unrelated persons from across the world. The stories are chilling, calling for urgent action before AI takes over. 

These stories look at everyone from underpaid gig workers to activists, refugees, single mothers, doctors, bureaucrats, and priests. Through these case stories, Murgia examines all the relevant facets of life, from livelihood to society and from freedom to what the future holds.

What comes out clear is that ‘surveillance capitalism’ is the denominator which guides the business model that monetises personal data. Amid all the hype and frenzy around AI, there is little denying that the prejudices of those who create it get amplified. For instance, facial recognition cameras are biased towards fair skin, and against religious minorities, migrants and refugees. Backed by data colonialism by tech giants, any scalable system under the capitalist economy is built to benefit large groups while excluding the lesser privileged. This can only trigger further exclusion and inequality.

An illusion of free will?

Each of the stories makes it clear that AI is out to compromise our agency and shatter our illusion of free will. Data workers are as vulnerable as factory workers, although, at present, they remain an undervalued bedrock of the AI industry. The algorithms that create deepfakes target women, who are hypersexualised by technology.

Facial recognition has empowered police to exercise nuanced judgement, promoting widespread human rights abuses.

In a world wherein AI promotes surveillance, censorship and control, freedom is first to get compromised with dissent becoming irrelevant. The stories in the volume are disturbing, exposing powerlessness and vulnerability in a world turbocharged by AI.

Generative AI has raced through the economy without anyone answering the compelling questions it has raised. Will the so-called knowledge workers still have work in a few years? Who owns the rights to all of humanity’s creative outputs? Will white-collar jobs continue to exist as they are? How can a society sustain itself without work?

The trouble is that people — scientists, economists and politicians — who are supposed to have answers to such questions are as much in the dark as the man on the street. Furthermore, in the post-truth era, preventing AI from producing false or biased content will remain a challenge.

Murgia is clear though that AI is indeed a way to augment human intelligence and solve impossible problems, but has better utility when it also preserves human dignity.

To protect human agency, the author has drawn a checklist of guiding principles to ensure that algorithms are not without ‘algor-ethics’, a basic framework of human values to be agreed upon by multiple stakeholders around the world, and implemented by machines. Written with empathy and deep concern about the future of mankind, the author raises questions about the unrestricted idea of selling back our dreams repackaged as the products of machines. Code Dependent makes for compelling reading as it concerns not just the impact of AI on our collective future but also its repercussions on humanity itself.