Open Intelligence: AI demands a rethink of human learning for digital age

Open Intelligence: Education Between 

Art and Artificial by Saikat Majumdar

Published by Penguin Random House

187 pages ₹499

 

The words of the 2024 Nobel Laureate in Physics, Geoffrey Hinton, feature in the epigraph of Saikat Majumdar’s Open Intelligence: Education Between Art and Artificial.  

 

Popularly known as “the Godfather of AI”, Dr Hinton notes that humanity “is just a passing phase in the evolution of intelligence”. This seemingly deceptive speculation is both profound and unsettling. Passing or not, this phase has taken a tumultuous turn as humans co-share this notion of intelligence increasingly with machines.

 

Since several tasks can now be outsourced to AI systems and tools, the question we must confront is what humans must learn to make them appear distinctly intelligent. This is an issue Mr Majumdar encountered in a boarding school in Ooty. In the book’s first chapter  titled “Art in the Artifice”, the creative writing teacher talks about “the paradoxes of education: its power to liberate and its tendency to confine”.

 

Invoking Yuval Noah Harari’s Nexus: A Brief History of Information Networks from the Stone Age to AI,  he seconds the concept that “intersubjective networks” shape human history, underlining how generative outputs by AI show signs of human biases.

 

Determining whether progress in AI development will reinforce this trend or create something novel is about as unpredictable as AI’s understanding and treatment of humans in the future. But recent history suggests that the consequences are inevitably directed at the most disenfranchised and marginalised. The “vitriol content directed at the Rohingyas to the top of then Facebook users’ newsfeeds, downplaying everything else” is an extreme example. And so are viral deepfakes, against which many public figures are seeking protection.

 

This chapter also underscores our tendency to confuse “novelty for originality”. A case in point is a test prompt this writer executed on Claude, by Anthropic. On being asked what’s that one thing humans can’t “comprehend fully” in the human-machine era, and the limits of understanding of AI-driven systems and humans, the chatbot concluded that “we are in the middle of a meeting between two kinds of intelligence that have never met before”.

 

This reviewer is unsure whether we are under- or oversubscribing to the “uniqueness” of human intelligence, owing to its “subjective consciousness”, as Mr Majumdar notes. Or if the evaluation of the “economies of the skills”—obsolete or upcoming — under the framework of this new reality is possible, as the author remarks.

 

Prima facie, what lies ahead is a challenge, which is growing and becoming more complicated at warp speed, which is why writing a book like Open Intelligence, which deals with current realities, is a risk. But there is an opportunity, too, to direct humans towards other ways of imagining AI-related concerns facing humanity. This is evident in the chapter “The Great Roulette Wheel”, in which Mr Majumdar reflects on humans’ preferred — or prescribed — way of learning in an educational system designed to make its beneficiaries become employees in future. That is, by “copying vast amounts of material from sources, committing enormous quantities to memory”.

 

Invoking the concept of “gambling”, the author argues that young people are losing “the happiest and most formative years of their lives in the gamble of examinations”. Must experience then be considered a barometer of learning, education, or intelligence, giving humans a distinctive edge? Can skills such as creative writing be taught? These are a few questions that the author makes the reader consider.

 

As it is with most small- or large-scale problems facing humanity, the future path for the human-AI nexus, too, is the middle one. The one involving their collective intelligence and shaping a collective future, which is a topic Mr Majumdar addresses in “The Tense Future of Intelligence”. The word play here won’t be lost on the seasoned reader. To speculate the future, one must inevitably turn to the past. The author does this, too, recollecting a time when people used to remember phone numbers by heart. Is it useful anymore? Maybe not. “The value of factual memory needs to be radically rethought in the digital age,” he notes.

 

Sharing his children’s approach towards language and problem-solving, the defeat of Lee Sedol, the world-famous Go champion, in a Go match by Google’s DeepMind, the “experiments in musical intelligence”, and the interplay of the concrete and the abstract in informing intelligence, Mr Majumdar concludes that emotions are but a “performance” of a kind.  They are “dependent on certain animate and inanimate patterns”, making one realise the role of “personalised consciousness behind” them. It is maybe this quality in Mr Majumdar’s view that makes humans distinctive.

 

Mr Majumdar appeals to the need to “imagine our education innovatively” to be “ready for an order of a very different kind”, which may perhaps enable us to tell our own story in ways AI cannot. That’s because, as he concludes, the “real danger of the loss of our humanity” isn’t “coming from the machines themselves. It’s coming from us”. 

 

(The reviewer is a Delhi-based writer and culture critic. @writerly_life)