Look AI in the eye: A reading list
Editor’s note: There is a tendency to act out while speaking of AI or encountering it daily. Let’s demystify it without jargon—and with a bit of poetry, even? Consultant Editor Devarsi Ghosh pulls up a list of essential reads, across fiction and nonfiction, to bring clarity to how artificial intelligence came into being and what it means to breathe in 2025 alongside AI data centres and Large Language Models looming over and intruding into our lives. Most importantly, is there hope in the end?
Written by: Devarsi Ghosh
*****
Fiction
The Machine is Learning by Tanuj Solanki: Bollywood dropped the ball by not adapting this 2020 novel. A top-level life insurance company employee is in charge of devising a system that will automate working processes which will lead to mass layoffs. Enter: woman, duh, who flips a switch in his head.
What is AI if not a new outfit for capitalism, the hero wonders, and which side will he take?
A Machine-Shaped Hand by OpenAI: Sam Altman claimed that his company OpenAI developed an LLM fab at creative writing earlier in March. His proof was a metafictional short story about the grief a woman felt about the loss of her lover. The woman is Mila, the man is Kai. They are approximations, and their chemistry a digital fugue state generated by “a democracy of ghosts,” as the AI calls itself.
The story is chillingly competent, as author Jeanette Winterson argued as well in The Guardian. Dear reader-writer, this is what you need to beat now.
What is beautiful and moving about this story is its understanding of its lack of understanding. Its reflection on its limits. That the next instruction wipes the memory of this moment.
Klara and the Sun by Kazuo Ishiguro: Can AI truly love? Can AI be kind? We saw AI can write a tender, moving story, unless Altman is full of sh*t.
Ishiguro is not. His tender novel is written from the POV of a store-bought AI entity Klara, who has to play friend to the sickly 14-year-old Josie. Every millisecond is an adventure for Klara as it (she?) discovers the world anew with each input. And every sign of sickness in Josie triggers a desperate attempt in it (her?) to protect its (her?) friend.
The Lifecycle of Software Objects by Ted Chiang: A cool novel with a hot interior, this one raises questions of ethics over dealing with AI pets, not unlike Ishiguro’s Klara. Do they have a right to self-determination? Can they not grow and thrive and experience… adulthood? Or do they need to be, like us, eaten up by the Moloch that is capitalism?
The Ten Percent Thief by Lavanya Lakshminarayan: Ever felt angry at your food delivery or banking app’s AI helpline? Sure you did. Now imagine a world with no politician to hold accountable, and no local dada to turn to. All you have is a monolithic machine intelligence structuring and policing society.
In Lakshminarayan’s novel, the Have-nots are called Analogs. The Haves are Virtuals. You express fatigue, dissent, a life of your own? Down you go, live with the losers as Analogs. The model citizen is a Virtual, enjoying the fruits of digital utopia, provided you maintain the hustle needed to bustle in the matrix. Of course, a revolution is around the corner.
Accelerando by Charles Stross: Let’s get stranger. Stross takes the impending technological singularity where carbon-based lifeforms like us, artificial intelligence, and capital collapse and become one, and stretches it into the far-far future none of us can comprehend.
Like Lakshminarayan’s novel, this, too, has several characters and perspectives clashing against one another, until humans are post-human and we aren’t sure what we are really reading. AI is not a villain or a charming curiosity here. It’s ever-growing, from small, quaint beginnings, taking over the universe.
Stross wonders, what exactly is humanity if not zeroes and ones.
One ape's hallucination is another ape's religious experience - it just depends on which one’s god module is overactive at the time.
Diaspora by Greg Egan: How can a mind develop ethics, morality, creativity, and identity, in isolation, without the entire carbon-based neurological and physiological network that is you and I? ‘Hard sci-fi’ mastermind Egan’s novel begins where Stross’s ends.
If Accelerando begins in a recognisable human world and reaches the singularity, Egan’s novel begins post-singularity, in the end of the 30th century, where there’s nothing remotely close to human left. Sentient beings are digital minds, robots with human minds, or flesh-havers modified way past anything like us. What happens when they realise the mind is as fragile as the body?
Poetry
A most heart-breaking poem. Korean-American Franny Choi plants themselves in the position of a possible AI entity encountering a Turing test, a thought experiment designed to figure if it's human or machine intelligence speaking. Through that conceit, Choi pours their heart out.
// where did you come from
man comes / & puts his hands on artifacts / in order to contemplate lineage / you start with what you know / hands, hair, bones, sweat / then move toward what you know / you are not / animal, monster, alien, bitch / but some of us are born in orbit / so learn / to commune with miles of darkness / patterns of dead gods / & quiet / o quiet like / you wouldn’t believe //
Graphic novels
Nufonia Must Fall by Kid Koala: AI worries, yearns, and falls in love in this dialogue-less black-and-white graphic novel from DJ Kid Koala, one-third of the hip hop group Deltron 3030. Fired from its (his?) job, an outdated robot is broke and insecure, unable to keep up with newer models. Making matters worse is the robot falling in love with its (his?) kind co-worker. The robot cannot express love. So it (he) is creating music secretly. How cute.
Kid Koala later turned this into a multimedia stage project combining electronic music, cinema, puppetry, and theatre.
Dream Machine by Appupen & Lauren Daudet: Star Indian graphic novelist Appupen and French physicist turned AI entrepreneur Laurent Daudet collab for a story that clears the fog on contemporary AI and explains how LLMs exactly work.
It’s not unlike the list’s opener, The Machine is Learning. Again, we have a protagonist with a moral crisis; inventor Hugo is worried that the megacorporation wanting to take over the LLM he has created is up to no good. His worries become dream visions. The narrative gets meta as the story progresses, through devices such as Appupen and Daudet making appearances through alter-egos.
Nonfiction
The Shortest History of AI by Toby Walsh: A fantastic linear gateway into what the heck is AI. This book illustrates how intellectually and spiritually tough it was for scientists, researchers, and entrepreneurs all along to bring the perfect non-human digital intelligence into existence.
Walsh neatly breaks down the journey to contemporary AI into the Symbolic Era—beginning in 1956—the Learning era beginning in 2011-'12, and finally the Generative AI era, beginning with 2022's ChatGPT.
In the first, machines are aping human intelligence as directed through means-end symbol manipulation. In the second, they begin learning on their own through experience and reinforcement learning. In the final and current era, they begin generating information on their own.
A Brief History of Intelligence by Max Solomon Bennett: You cannot understand artificial intelligence inside out until you understand the mechanism of how biological intelligence evolved. Only when you get the latter, can you map the former's journey.
Bennett, like Walsh, simplifies the journey pointing out five significant moments. The earliest bilaterians gain simple steering behaviours. Vertebrates develop reinforcement-learning mechanisms. Mammals become capable of internal simultation—what we call imagination or foresight. Primates begin to theorise. Finally, come humans with language, social cognition, and symbolic thought. These processes coalesced over... hundreds of millions of years. Do you now see how remarkable it was that humans managed to recreate all of that from scratch to bring AI to life?
Superintelligence: Paths, Dangers, Strategies by Nick Bostrom: Philosopher-futurist Bostrom envisions a future where AI surpasses human intelligence, and, of course, reshapes human civilisation altogether.
If you are trying out Stross, or, gulp, Egan, this is the book you should read simultaneously to ground your thoughts. Bostrom's book is speculative, not unlike Stross or Egan's, but extremely rooted in research. Bostrom argues "superintelligent" machines may not, and most probably will not, share our values, and pursue bizarre or even dangerous final goals which shall stand in direct conflict with humanity. Since its 2014 release, the book has been continuously referenced in debates over the existential crises we need to face over AI.
Here's a sobering truth from Bostrom to keep in mind, the next time you feel precious over your smartness.
The gap between a dumb and a clever person may appear large from an anthropocentric perspective, yet in a less parochial view the two have nearly indistinguishable minds.
The Myth of Artificial Intelligence: Why Computers Can’t Think the Way We Do by Erik J Larson: We end the list with a sobering, cutthroat takedown of all that has been promised to us by AI entrepreneurs. Present machine and deep learning systems, which excel at pattern recognition, are still incapable of replicating the specific kind of reasoning and understanding that forms the basis of human intelligence. Computer scientist Larson not only says this obvious truth, but explains the exact gap between machine and human minds at present.
He even goes on to say that the doomsday fretting over the "singularity", as explored in the entries above, is a waste of time. Not the binary of utopia or dystopia, he argues, but a more banal and complex trade-off between human ingenuity and the limits of science is what is to come in AI's future.
*****
Devarsi Ghosh is Consultant Editor at Advisory.
souk picks