Can AI write decent book reviews?

Or: can AI write decently at all?

As ChatGPT is on everybody’s lips at the moment, I thought I’d revisit an article I wrote several years ago, and amend it by bringing in book reviews.

Can AI write?

The BBC used artificial intelligence to generate reports on the UK General Election results. The AI “wrote” an article for most of the Constituencies when the results were in, apart from a few. Human editors checked the articles and sometimes added a bit of context.

Gary Henderson, a teacher and Director of IT in a UK school, asks a very interesting question:

Using AI to generate the news articles you read.
Does it make a difference whether it is a human or an AI writing the news we read?#DigitalCitizenshiphttps://t.co/ubwcssG7pP pic.twitter.com/23QeR5yUEE

— gary henderson (@garyhenderson18) December 15, 2019

I think it does make a difference, for several reasons.

First, these kind of articles can at present only create articles based on data. The statistics are presented to the reader, and that is pretty much that. As far as I can see, each article has been written in the same way, with the facts presented in the same order. For example, the name of the winner is given, followed by the number of votes by which they won.

But how does the AI know to do that, and why are some data cited and not others? For example, in Kensington, the Conservatives won by 150 seats. But how many people voted for them altogether?

Obviously, human writers will make a decision, or use a style guide, to dictate the answers to questions such as these. My concern is that we can easily hold people to account should they present the data in a way that misleads people. But how do you question an algorithm once (a) it has been established and embedded enough that nobody really knows how it works? And/or (b) that is (probably) not in the public domain for intellectual property reasons?

The headlines sometimes also assume prior knowledge which is not always addressed in the article. For example, the headline for Ilford South reads:

Ilford South: Mike Gapes Loses Seat

That’s quite anodyne. What it does not tell you is that Mike Gapes lost his seat after 27 years, because he left the Labour Party. Unfortunately for him and others, people in the UK tend to vote for a party (or against a party) rather than a person.

So that headline is very objective, because it hides much background data (Gapes left the Labour Party after 27 years: why? You used to be able to read the answer on his website, but now his website appears to be inactive.) However, because the headline and the report omit important background information, you could argue that it’s not objective, in the sense of being impartial, a point made (in a different context, obviously) by Christopher Goering and Paul Thomas (Eds) in Critical Media Literacy and Fake News in Post-Truth America (Amazon affiliate link).

Another issue is that these articles are mind-numbingly boring, which I’m actually rather pleased about. Why? Because as long as AI can’t write creatively, people like myself who earn money from writing have little to fear.

Indeed, I think a partnership could work really well. In the context of education technology (the area in which I write the most), I’d love to have a bot that could trawl through online archives, pull out some great stats and then write a report presenting the data. I could then analyse that data, adding colour, context and nuance to the basic report generated by the AI.

Getting back to Gary Henderson’s question, I suppose the answer is as follows. If you just want the basic facts, the bare bones, then perhaps it doesn’t make a difference. The AI can do a good job of reporting election results, football scores and weather patterns. You don’t need writing skills to be able to do that, only research skills. So ultimately, if the AI can’t get the data for itself (if it isn’t available online, say), then all you need is some low-paid research assistants to obtain the data and feed it in. It could lead to few jobs for journalists and more opportunities for interns.

It’s amusing and interesting to note that Jonathan Swift made the point that writing skills are not needed if you have a writing machine nearly 300 years ago, in Gulliver’s Travels (1726). Having met a professor who has invented a writing machine he says:

Everyone knew how laborious the usual Method is of attaining to Arts and Sciences; whereby by his Contrivance, the most ignorant Person at a reasonable Charge, and with a little bodily Labour, may write books in Philosophy, Poetry, Politics, Law, Mathematics and Theology, without the least assistance from Genius or Study.

However, if you want reporting that can bring the data alive, place it in a wider context, and evince an emotional response or connection from the reader, then it does make a difference. If you don’t believe me, you just have to compare the late sports writer Hugh McIlvanney’s depiction of cold weather with a factual report. The latter might have stated something like:

The temperature was 2 degrees Celcius but because of the wind felt like minus 6 degrees.

In McIlvanney’s hands, that sort of thing was transformed into:

It was the kind of wind that stripped the flesh off your bones, then came back for the marrow.

The difference is that in the latter you can almost feel the cold. And that to me is the difference between an AI-written article and a human-written one. In fact, to put it another way, I think Arthur C. Clarke’s observation about teachers vis-a-vis teaching machines may be applied here, namely:

In my opinion, any writer that could be replaced by AI probably should be.

Book reviewing

I thought I would test ChatGPT’s mettle in a rather self-interested way. I write a lot of book reviews — a lot. I have three books I need to review for an education magazine by 21st January, plus two books I need to review for other websites soonish, and I have to write a review of an exhibition for a different education magazine by next week <gulp>.

Therefore I have two (competing) concerns.

Firstly, although writing book reviews is not tantamount to owning a money-printing press, it is quite enjoyable, helps me keep up-to-date, and saves me having to spend a small fortune on books. But if ChatGPT can write better reviews than I can, why should an editor pay me?

Secondly, if it can write better reviews than I can, why don’t I save myself a lot of time and just get AI to review the books without my even bothering to read them? Not that I would: I’m rather more honest than that, but you can see where I’m coming from can’t you? Could AI, in other words, be used to game the system?

To test these ideas out, I asked ChatGPT how to write an interesting book review. Here’s what it said:

How Can I Write An Interesting Book Review

Writing an interesting book review is a great way to share your thoughts on a book with others and to encourage them to read it. Here are some tips for writing an interesting book review:

Start with a brief summary of the book: In a few sentences, summarize the main plot, characters, and themes of the book.

Explain why you enjoyed or didn't enjoy the book: Use specific examples from the book to illustrate your points. Did you like the characters? Did the plot keep you engaged? Did the writing style work for you?

Evaluate the book: Was the book well-written? Was it well-researched? Did it make you think or feel something?

Consider the intended audience: Is the book aimed at a specific age group or type of reader? How well do you think it will appeal to that audience?

Use quotes sparingly: Including a few quotes from the book can help illustrate your points, but be sure to balance them with your own thoughts and analysis.

Conclude with a recommendation: In the end, it's up to you to decide whether or not you would recommend the book to others. Explain why you would or wouldn't recommend it, and to whom.

Overall, the key to writing an interesting book review is to be honest, thoughtful, and specific. By sharing your own thoughts and experiences with the book, you can help others decide if they would like to read it, too.

A couple of things stand out for me from that advice.

Firstly, it seems very bland to me. I don’t know: perhaps I’ve been writing book reviews for so long that I take a lot of this sort of thing for granted. I mean, it’s not bad advice, but somehow very generic.

Secondly, the first paragraph says that writing a book review is a great way of encouraging others to read it. But isn’t that making a big assumption? One of the reasons I write book reviews is to be able to advise teachers or writers (my main audiences) on whether or not they should read the book. Sometimes my conclusion might be: don’t waste your time or money.

I then asked it to write a review of The Old Man and the Sea. I’ve never read that book, so I don’t know how accurate the review is. All I can say though is that I don’t have much of an idea of what the book is about. Nor are there any quotes given. Thus the AI has broken two of its own rules of book reviewing.

To be fair, I’m afraid I can’t always obey all these rules either, because I’m given only 150 words for the reviews in one magazine. But I didn’t impose any such limit on ChatGPT, so I should have thought it might have been useful to illustrate how Hemingway brings his characters alive. Anyway, read the review for yourself and see what you think.

Review Of The Old Man And The Sea, by ChatGPT

I recently finished reading "The Old Man and the Sea" by Ernest Hemingway, and I must say that it was a beautiful and powerful read. The story follows an elderly Cuban fisherman named Santiago as he embarks on a journey to catch a giant marlin.

Hemingway's writing is, as always, impeccable. His prose is lean and evocative, and he manages to convey a sense of timelessness and universal themes despite the seemingly simple plot. The descriptions of the sea and the struggle between Santiago and the marlin are both vivid and lyrical, and the character of Santiago is deeply moving and relatable.

One of the things I love about Hemingway's writing is the way he is able to convey so much with so few words. He doesn't waste any time with unnecessary exposition or florid prose, and yet he still manages to create fully realized characters and landscapes. "The Old Man and the Sea" is a testament to Hemingway's mastery of his craft, and it is a must-read for any lover of literature.

Overall, I highly recommend "The Old Man and the Sea" to anyone looking for a poignant and beautifully written tale of resilience and determination. It is a true classic that will stay with you long after you finish reading it.

Two further reviews

I also asked ChatGPT to review two books that I’ve already reviewed. I’m not going to ask you to guess which of the following reviews were written by the AI, because other people have done that sort of thing to death already. But I thought you might find it interesting to compare the reviews side by side as it were. See what you think:

Review of Blurb Your Enthusiasm, by ChatGPT

Blurb Your Enthusiasm: An A-Z of Literary Persuasion is a clever and insightful guide to crafting compelling book blurbs. Written by Louise Willder, an experienced copywriter and book editor, this book is filled with practical tips, examples, and exercises to help writers and marketers create captivating blurbs that will entice readers to pick up their books.

One of the things that sets this book apart is its comprehensive coverage of the topic. Willder covers everything from understanding the purpose of a blurb and identifying your target audience, to crafting attention-grabbing headlines and avoiding common pitfalls. She also provides a wealth of examples from various genres, demonstrating how to effectively apply the concepts she covers.

In addition to being informative and useful, Blurb Your Enthusiasm is also a highly enjoyable read. Willder's writing is engaging and entertaining, making the book a pleasure to read from cover to cover.

Overall, Blurb Your Enthusiasm: An A-Z of Literary Persuasion is an excellent resource for writers and marketers looking to improve their blurb-writing skills. It's comprehensive, practical, and enjoyable, and it's sure to help you craft blurbs that will capture readers' attention and persuade them to give your book a chance.

Review of Blurb Your Enthusiasm, by me

Your challenge, should you choose to accept it, is to write approximately one hundred words for a book cover flap. Your blurb (as it’s known in the UK) must summarise the theme, entice potential buyers, but not appear to be written by the author’s mother.

Take into account the avoidance of hackney-phrases, inadvertent sexism and words that simply don’t sell, and you can see that it’s much harder than you might imagine. And you can also see that blurb-writing involves skills that English students must develop, such as concision, awareness of audience and writing for different purposes.

Given that blurbs are a form of marketing, this would suit Business Studies students too.

Or simply read it for pleasure: it’s well-written, humorous, and very informative. It includes examples of good and bad writing, and provides an historical perspective. A must-have addition to your class or school library.

Review of How to be a writer, by ChatGPT

How to be a Writer, by Marcus Berkmann, is a humorous and relatable guide to the ups and downs of the writing life. Through a series of witty essays, Berkmann offers insight and advice on everything from dealing with writer's block and tackling self-doubt, to finding inspiration and navigating the publishing world.

One of the things that sets this book apart is its honesty and authenticity. Rather than presenting a glossy, romanticized version of the writing life, Berkmann offers a candid and realistic portrayal of the challenges and frustrations that writers often face. His humor and self-deprecating tone make the book a pleasure to read, and his insights are valuable and thought-provoking.

Overall, How to be a Writer is a great resource for writers of all levels, whether you're just starting out or have been writing for years. It's funny, wise, and full of practical advice, and it's sure to provide comfort and inspiration to anyone who has ever struggled with the writing process. So, it is a must-read for all writers.

Review of How to be a writer, by me

The life of a freelance writer is one of unreliable remuneration, shifting loyalties and sudden endings, as this book makes abundantly clear. Written with humour and candour, it's a welcome dose of reality amid the plethora of books and blogposts promising riches beyond measure to anyone who can wield a pen. As Berkmann makes clear, success is dependent on luck as well as talent, and can be precarious.

He is, however, perhaps too dismissive of his own abilities - even the pieces he cites as bad examples of his writing are pretty good.

Arguably more memoir than guide, it features no index and the chapter titles give little hint as to their contents. All the same, lend a copy to anyone who romanticises about the writing life - it may not deter them, but at least their eyes will be more open.

Concluding remarks

OK, I know I may be slightly biased, but I think my reviews are better. They have a bit more personality, and they read as if I actually have read the books, which I have. (If you disagree, and think that ChatGPT’s reviews are more pleasurable to read, you need not feel obliged to say so in the comments section!)

But could ChatGPT be useful at all in the context of book reviewing?

I can think of a couple of uses. One is that you could use it to generate reviews of books you haven’t read in order to fill up space in a newsletter or to save time. For example, I have another newsletter, a books bulletin, and I suppose I could make it bigger with more variety if I employed ChatAI to do some of the work for me. After all, the reviews are passable. They’re not going to set the literary world alight, but they’re reasonably accurate because, presumably, ChatGPT has been fed reviews of these books in order to learn about them. I won’t be doing that myself any time soon, but it’s a possibility.

I wondered if another use might be to obtain an evaluation of books you haven’t come across in the reviews pages (most books don’t get reviewed). However, when I tried this I received this message:

Without more context, I am unable to provide a review or any other information about the book. If you have specific questions or topics you would like to discuss, I would be happy to try to help. Please let me know how I can assist.

Which is good news really isn’t it? We humans are useful after all.