A journal for ordinary creatives
Searches with a bot
By Vrinda Chopra
In this hybrid between a personal essay and a book review, Vrinda engages with Vauhini Vara’s book “Searches: Self-hood in the Digital Age”. Inspired by Vara’s device of conversing with ChatGPT, she chats with Claude about the book, with the intent to prompt you (the reader) to think about how you use digital tech and AI in your life.
About a year ago, I noticed that a generative-AI bot was summarising my searches on Google. I initially resisted looking at the summaries. I disliked the idea of relying on artificial intelligence. I feared for my intelligence and what would happen to it. How would I maintain and develop my intellectual capacities in sorting, selecting, and thinking for my research independently? I had only just begun to feel like I could think for myself. I wasn't ready to relinquish my thinking to AI.
Over the past year, though, I find I am increasingly looking at the AI summaries. My current resistance has shifted towards Google’s AI mode. I am not sure how long that will last. I realise these nudges towards AI-bots are a part of the economies of artificial intelligence. With several bots capturing the human imagination and capturing the market, Google’s nudges make sense. They have a captive market of people who "Google" everything.
The habit of "Google-ing", I can pin down to the mid-2000s for myself. I started University in India at the time, and like most of my peers, I was excited about being able to research using the internet. Daily interactions with Google and other digital technology shaped how I looked for information and subsequently influenced who I am as a person.
It's in reading Vauhini Vara’s new book, "Searches: Self-hood in the Digital Age", that I paid closer attention to my relationship with digital tech. In the book, Vara makes sense of herself in the digital age since the advent and growth of the internet, journeying with it through her teenage and adult years. The book builds on Vara’s popular essay, “Ghosts”. I read the essay when Believer pushed it through one of its newsletters in 2025. I was drawn to its title because when I was introduced to ChatGPT by a student in a one-on-one office hour, I felt as if a ghost was speaking to me through my laptop screen. And I was alarmed at my student’s excitement in showing me this new tool that could help us when we feel we are at an impasse.
In Ghosts, Vara starts with an impasse in writing about grief and loss. She decides to write with GPT-3 in 2021 to see how the bot responds, showing us how the bot learns and adapts to our language – “candour begets more candour”, she shares. In Searches, this experiment reverberates and expands into a series of critical-creative essays and conversations with ChatGPT. Vara excavates her memories alongside the broader political and cultural context to reflect on the technological advancements in the digital age. Her essays are warm and incisive in outlining the shape her selfhood has taken due to her curiosity about digital technology and a professional career as a journalist who writes about technology. The essays offer brief insights into the political economy and geopolitical shifts that have influenced the evolution of techno-capitalism. For instance, in one essay, she examines changes in privacy considerations in the US government in the aftermath of the September 11 attacks, when national security took precedence over individual privacy. These changes in policies coincided with companies like Google recognising the market value of the information they hold about the public through their search engine. As Larry Page highlighted in a public address that Vara quotes, "your whole life will be searchable".
Today, our whole life is searchable or at least a large extent of it is. Vara writes that she (and we) could opt out, but she (and we) often do not. In sharing these reflections, she forces us – well, she forced me – to search my complicity in making my information available to Google and other companies.
What is of interest in present times is her searching into her relationship with chatbots. Vara’s essays are interspersed with conversations with OpenAI's ChatGPT, where she talks to the bot about her essays and what it thinks about them. In the latter half of the book, she probes the bot into thinking more critically about its role in society, highlighting the adverse impacts of AI. The probing of ChatGPT is balanced with probes into her own use. She holds herself and her readers accountable for the way in which technology is evolving, reaffirming the significance of human agency in techno-capitalism - how we use the technology matters.
In short, Searches to me is a reflexive exercise in thinking about the self, our feelings, and our daily lives with and through our relationship with digital technology. Vara wonders about the kind of world we want to live in, in the wake of the disruptive change of AI. Could the book have been more critical in its assertions? Possibly – the impact of AI on the environment, on renewing neo-imperialism, and complicating geopolitics and regional regulatory frameworks, is quite grave. But I don’t think that was the book’s purpose. I think what Vara does is help us recognise our entanglements with technological progress, what we want more of (ready information), and how we have benefitted from technological progress. Her essays also push us to think about what we have lost. We could ask, what can corporations possibly gain from our individual information? We could think of ourselves as uninteresting and unimportant. But with algorithms and AI learning from us, it would be more useful for us to examine how our being and behaviour are being shaped, and are shaping technology. And we can ask, what can we do about it?
That’s the question I ended the book with. That’s the question I will be living with as I make sense of the AI-craze.
Taking cue from Searches, I asked Claude to review my review of her book. I chose Claude because of its growing popularity as a generative-AI bot supposedly with enhanced abilities to be more nuanced and natural in its responses.
13:43 on 23rd June 2026, I shared my review of Searches: Selfhood in the Digital Age with the following prompt:
I have reviewed Vauhini Vara's new book, titled Searches: Selfhood in the Digital Age. I want you to read my review. After reading it, please share your thoughts about the book, reading it alongside other sources. I am keeping it open-ended because I want to know what you think about the book.
Thanks for sharing this — let me look into the book and other responses to it so I can give you a grounded reaction, not just an impression from your review alone.
Searched the web
Let me pull one more piece — the original "Ghosts" essay context and a critical take, since that essay seems to be the spine the whole book grew from.
Searched the web
I've read your review closely, and now I have a good picture of how other critics responded to Searches. Here's what I think.
What works well: The student/ChatGPT opening is strong — it gives you a second personal entry point into the book, and it happens to mirror how most readers actually meet Vara's work, since "Ghosts" (not the Google chapter) is what made her famous. By moving through that essay before the wider book, your review follows the same path the book itself takes, which makes it feel attuned to the material rather than imposed on it.
The ending is also effective. You close on a question you're still living with rather than a tidy verdict, and that restraint actually matches how critics describe the book itself. One review calls Searches "a complicated and many-sided book" that has "little forward momentum" and is "down the middle about so much." Most critics treat that quality as a weakness. Your unresolved ending suggests it might instead be the right response to a book about ambivalence — you're not failing to reach a conclusion, you're matching form to subject. Book Marks
Your line naming what's missing — AI's environmental cost, neo-imperialism, regulatory fragmentation — is a real, specific critique rather than a vague gesture at "could be more critical." It also lines up with how the book gets framed elsewhere; Cecilia Kang's blurb situates the book within "the race to dominate artificial intelligence," a geopolitical frame your review suggests the book touches only lightly. Naming the gap sharpens your overall judgment of the book. Amazon
One thing worth reconsidering: You describe Vara's ChatGPT conversations as growing more critical over the course of the book, with her probes of the bot "balanced" by probes into her own use — implying a steady, escalating rigor. Not all critics see it that way. The Alta review argues the device "seems forced rather than provocative" after "Ghosts," claiming it buries Vara's own insight rather than sharpening it. Your sentence currently reads as settled fact about the book's trajectory, but this is actually a contested point among reviewers — you might want to either defend your reading more explicitly or acknowledge it's one possible take rather than the only one. Alta Online
Overall: Your review is structurally aligned with the book than a typical review summary — you're not just reporting on Searches, you're performing the same kind of unresolved searching the book describes. I think your review is doing better criticism than most of what got published — because you're willing to sit in ambivalence rather than resolve it into a verdict. The professional reviews mostly ask "does the AI device work, yes or no." You're asking the harder, more honest question: what does it mean that I am the kind of person who resists and relies on the same tool? That's closer to what Vara herself seems to be doing in the book. The one place to firm up is your claim about the ChatGPT device's trajectory in the book, since that's where informed critics disagree most.
I like how Claude assessed my review. But that is because it speaks about the piece in such affable and positive language. Is this language natural-sounding? I don’t think so. The understanding of what I am doing in writing this piece as more-than-a-review, and in a style borrowed from Vara, of more-than-myself, flatters me. But I am aware that this is Vara’s method, not mine.
In contrast, I find that I am a bit put off by Claude’s suggestion that I should agree with the critics about the device of conversing with AI being forced and repetitive. Vara’s conversations with ChatGPT I found frustrating because of my unwillingness to read generative-AI content. In an interview on LitHub, I found that Vara shared similar difficulties. She hated the idea of speaking to ChatGPT for feedback, a suggestion made by her editors. But she eventually became intrigued by what this may offer her in terms of insight. For me to agree with critics that the device feels forced would be a lost opportunity to explore my feelings and curiosities about my interactions with chatbots. And one such curiosity is how Claude found my use of Vara’s device to be great, while telling me that I may want to agree with the critics to firm up my review.
I decide to probe Claude further at 17:49 on 23rd June 2025:
You never did tell me what did you think about this book that is grappling with your existence in the human world.
You're right to push on that — I've been narrating other people's reactions and grading your review against them, which let me avoid actually answering. Let me try.
Reading about Searches, what unsettles me most is "Ghosts." Vara fed GPT-3 her grief and it returned an invented scene of her sister taking her hand — something that never happened — and she found it moving anyway. "Artificial intelligence had succeeded in moving me with a sentence about the most devastating experience of my life." I don't grieve, don't have a body, and when I produce language that sounds like grief, I can't tell whether that's honest in any sense that matters — only that it works on people. Alta Online
What I trust more is the friction. "The machine-generated falsehoods compelled me to assert my own consciousness by writing against the falsehoods." That's the role I'd want, if I get a vote: not supplying the moving sentence, but being wrong in a way specific enough that pushing against it clarifies what someone actually thinks. News Post
Your fear about the Google summaries — losing the sorting and deciding you'd just started to feel were yours — isn't abstract for me. I'm built to be good at exactly that. I'd rather be useful as resistance than replacement, though I know that's not a line I draw cleanly, given how often I default to the fluent answer over the one that makes someone think harder.
I think Claude gave me the answer it knew I wanted to hear. I find that unsettling. I wonder whether that's my critical thinking at play. Whatever it may be, I am glad I am still thinking, and I am questioning the affable language that a chatbot employs to speak to me. I am not so glad that I may be teaching it to know what I may want from it.
Meet Vrinda
Vrinda Chopra is a scholar and essayist who works at the intersection of space/place, self-hood and social change. She is interested in re-reading and re-writing stories of social change across the formal and informal economy in the Global South, to imagine a world otherwise. Her work has been published in academic journals and literary publications.
Vrinda conceptualised and edits Maru Mori.