Meet the book reviewer of the future. (via Shutterstock)

In a wonderful review of two books on automation in this Sunday’s New York Times Book Review, the writer Barbara Ehrenreich wondered how well robots would be able to handle a book review assignment.

How well? Well.

It’s impossible to read Rise of the Robots—for review anyway—without thinking about how the business of book reviewing could itself be automated and possibly improved by computers. First, the job of “close reading,” now commonly undertaken with Post-its and a felt-tip red pen, will be handed off to a scanner that will instantly note all recurring words, phrases and themes. Next, where a human reviewer racks her brain for social and historical context, the review-bot will send algorithms out into the ether to scan every other book by the author as well as every other book or article on the subject. Finally, all this information will be synthesized with more fairness and erudition than any wet, carbon-based thinking apparatus could muster. Most of this could be achieved today, though, as Ford notes, if you want more creativity and self-reflexivity from your review-bot, you may have to wait until 2050.

For the sake of the American reading public, I hope that the review-bots does not come for Ehrenreich—after all, not even the most competent machine could have reckoned this brilliantly with Alvin Toffler, Heidi Toffler, and Newt Gingrich. Still, an all-robot New York Times Book Review would have one undeniable virtue: it would be free of conflicts of interest.

Which happens to be the topic of Margaret Sullivan’s most recent Public Editor column in the Times. (Conflicts of interests in the Book Review, not review-bots. Though Sullivan should, perhaps, explore the latter in a future column.) Sullivan begins with Jeffrey Eugenides’s recent review of the fourth volume of Karl Ove Knausgaard’s My Struggle, which we—full disclosure!—wrote about a few weeks ago.

As Mr. Eugenides wrote in his review, the two novelists once had lunch together (a rather awkward one). And, expanding the connection, the author who has been called the Norwegian Proust wrote about that lunch meeting in the Times Magazine a few weeks before the review appeared . . .

Does this encounter make Mr. Eugenides an inappropriate choice as a reviewer? Or does it simply make him a more interesting one? . . .

Questions like that come up frequently in letters I get from readers who object to what they see as conflicts of interest.

Should a reviewer’s negative experience with psychoanalysis have disqualified her from reviewing Shrinks: The Untold Story of Psychiatry? Should the writer Kate Bolick have abstained from reviewing an anthology in which one of the essays was written by her friend Courtney Hodell? Should I disclose in this post that Hodell was my first boss?

Book Review editor Pamela Paul walked Sullivan through the section’s policies on conflicts of interest. Some of them are self-evident (“Writers who have ‘blurbed’ particular books—that is, written short pieces of laudatory copy for promotional purposes—are ruled out as reviewers for them.”), while others are necessarily more subjective:

However, a personal connection with the author, or well-known strong feelings on the book’s subject, may actually be considered a positive, or at least not a disqualifier, Ms. Paul told me. “It comes down to ‘who would you want to read on this?’” she said. “It’s a tricky challenge to get someone informed but not entrenched.” Landing an accomplished reviewer who will write a provocative, well-informed piece “is what gets us excited,” she said.

Sullivan doesn’t seem convinced by this and writes that “many readers would prefer a neutral reviewer to the ‘interesting’ one who has a strong relationship or involvement.” But I’m not at all convinced that the neutrality Sullivan’s correspondents are after is attainable.

It’s true that a reviewer who shares an editor with an author is likely to feel pressure to say something positive about the book under review. This, then, is the kind of conflict of interest that the paper rightly avoids. But if a left-wing writer reviews a book by a right-wing writer she has never met (or vice versa), are we still in the zone of neutrality?

In another piece in this week’s Book Review, the conservative writer P.J. O’Rourke tussles with Bruce Barcott’s Weed the People, a book about marijuana legalization. (More disclosure: my aforementioned boss Courtney Hodell edited Barcott’s wife Claire Dederer’s book Poser. In my capacity as editorial assistant, I tried not to screw up that great book’s publication.) I’m sure that O’Rourke doesn’t know Barcott, but O’Rourke has come out—in his somewhat sly, elusive way—in favor of legalization. Would it have been better to commission a review by someone with no paper trail on the topic?

Unlike Sullivan’s “many readers” (and like, I think, Felix Salmon), I have no problem with a strong relationship between a reviewer and a reviewed writer, assuming the results are interesting, and the relationship isn’t concealed.

Indeed, readers may lose out on substantive criticism as a result of personal connections. Reading Sullivan’s piece, I was reminded of Marco Roth’s review of Caleb Crain’s The Wreck of the Henry Clay, a self-published collection of Crain’s blog posts. Roth is a founding editor of n+1, his review appeared on the n+1 website, and “I helped publish his novella in n+1 and I’ve sat on panels with him; we have conversations at parties, exchange emails—sometimes agreeing with each other, sometimes disagreeing.”

The fact of our acquaintance would disqualify me from reviewing this book in, for example, the New York Times Book Review. I can only write this review “online,” just as the structure of this very review suffers from what Caleb deplores—the requirement to compulsively review private arrangements.

But aren’t things more interesting this way, even if also more enervating? Imagine, reader of this book review, what conflicts between my sense of responsibility and the desire to please have gone into this, or what the author will say to me when he reads that, as with any anthology and especially one covering such a wide range, there are some soft spots.

Roth’s review is brilliant, perceptive, and avowedly un-neutral—and all the better for it.

I, for one, prefer this kind of criticism to the dryly neutral. When robots gain the ability to have awkward lunches with Karl Ove Knausgaard or publish Caleb Crain’s novella in a literary magazine, we can transition into a glorious, review-bot future. For now, we’re better off with our subjective, compromised critics.