Algorithm Fatigue Is Changing How Readers Want to Find Books

Algorithm fatigue is often described as an irritation with endless scrolling, repeated recommendations and platforms that seem to know what someone will click next. But for readers, the deeper frustration is more specific: discovering a book can start to feel less like finding something meaningful and more like being processed by a machine. The question is not whether recommendations are useful. It is what readers want from discovery when prediction begins to crowd out curiosity.

There is a particular kind of tiredness that comes from opening a social platform looking for something new and finding a carefully arranged version of what you have already seen. The names may change. The covers may be different. The wording may be fresh. Yet the mood, genre, popularity signals and assumptions about your interests remain stubbornly familiar.

For readers, this creates an awkward contradiction. People often want help choosing their next book. Few have unlimited time to browse, and there are more books available than any individual could reasonably investigate. Recommendations can be genuinely useful. The problem begins when discovery becomes so dependent on prediction that the reader loses a sense of participation in the choice.

Algorithm fatigue is therefore not necessarily a demand for a world without algorithms. It is a demand for a different relationship with recommendation: less passive consumption, more context; less repetition, more surprise; less pressure to react immediately, more room to decide.

The real fatigue is not too many books

Readers are not tired because there are too many books in existence. Abundance is part of the pleasure of reading culture. A large catalogue means there are always forgotten novels, strange combinations of interests, small presses, debut authors and older works waiting to be rediscovered.

What exhausts people is the work of sorting through abundance when the sorting system is opaque. A recommendation may appear because a reader finished a particular title, because many people interacted with it, because it is commercially prominent or because it resembles a previous click. The reader may not know which of those things is happening. They are asked to trust the result without being given much of a reason.

That matters more with books than it does with many short-form posts. A book asks for time, attention and a degree of emotional commitment. Choosing one is not a tiny decision made between interruptions. It may mean several evenings, a long journey, or an attempt to find a particular feeling at a particular moment.

A recommendation that says, in effect, “people like you engaged with this” is less helpful than one that explains why the book might matter. Is it sharply observed? Comforting? Unsettling? Slow but rewarding? A good choice for someone who wants intricate plotting rather than lyrical prose? Readers often need judgement and context, not simply more items placed in front of them.

Why algorithmic discovery can make reading feel performative

Social platforms do not only help people find things. They also teach people what appears to be visible, valued and worth responding to. Over time, readers can begin to encounter books through signals of performance: what is trending, what is being discussed most loudly, what generates a recognisable reaction, or what can be presented quickly as a recommendation.

None of this makes popular books unworthy. Nor does it mean that enthusiastic online discussion is artificial. The difficulty is that a feed can make every discovery feel as though it arrives with an implied task. Read this because everyone is talking about it. React in the approved way. Post before the conversation moves on. Keep up with the list.

That is a very different experience from choosing a book because a friend described its atmosphere, because a passage stayed with them, or because a conversation revealed an unexpected connection. The latter forms of discovery may be slower, but they give the reader a role. They allow someone to say: this is what I am looking for, this is what I noticed, and this is why I think it might suit you.

For readers already weary of constant online performance, book culture can become another place where enjoyment is measured through visibility. A novel is no longer only something to read. It becomes something to keep up with, rank, defend, photograph, review or use as evidence of taste. Conversation can enrich reading, but it can also make private enjoyment feel insufficient.

What readers may want instead of a better guess

The obvious response to disappointing recommendations is to ask for a more accurate algorithm. Sometimes that is sensible. Better personalisation may reduce irrelevant results. But accuracy is not the only measure of a good discovery experience. A system can become very good at predicting what a reader has previously chosen while becoming worse at helping them encounter anything unfamiliar.

Readers may want four things that prediction alone cannot supply.

• Context. They want to know why a book is being suggested, what kind of experience it offers and what other readers noticed about it.
• Agency. They want to express a mood, question or curiosity rather than be treated as the sum of previous behaviour.
• Serendipity. They want room for an unexpected title that does not fit neatly into their existing reading history.
• Conversation. They want recommendations that come from people with a point of view, not only from engagement patterns.

These needs point towards a more human form of discovery. It does not have to reject technology or pretend that every personal recommendation is wise. It simply recognises that readers often choose books through interpretation. A recommendation becomes more valuable when another person has done some of the thinking in public: explaining the appeal, naming the caveat, comparing it with something else, or admitting who might not enjoy it.

Conversation is a different kind of recommendation engine

A conversation about books does not behave like a ranked list. It can wander. Someone may ask for a novel with a particular emotional quality rather than a genre label. Another person may recommend a book they disliked but found impossible to forget. A reader may discover that the thing they want is not “historical fiction” but a story about family loyalty, moral compromise or starting again in an unfamiliar place.

This kind of exchange is valuable because it makes the reasons behind a recommendation visible. It also leaves space for disagreement. One reader’s favourite may be another reader’s overhyped disappointment, and that difference is useful information rather than a failure of the system.

Bobble is designed around this book conversation: books, authors, genres, recommendations, reactions, discoveries, opinions and what readers loved or hated. Its significance is not that it promises to remove the difficulty of choosing. No community can eliminate uncertainty, and uncertainty is part of what makes discovery interesting. Its value is in treating the discussion around books as part of discovery rather than as noise surrounding it.

That distinction matters in an era of algorithm fatigue. A focused space for book conversation can give readers somewhere to ask a more precise question than “what is popular?” They can ask what to read after a particular kind of disappointment, whether a supposedly difficult book is worth the effort, or which author handles a subject with unusual care. The answer may still be subjective. That is precisely why it can be useful.

Readers who want that kind of exchange can explore Bobble, The Globe’s space for book and reader conversation.

Focus can be more valuable than infinite choice

One reason general social feeds become tiring is that they ask users to move rapidly between unrelated forms of attention. A book recommendation sits beside a joke, an argument, a news item, an advertisement and a personal update. The reader is expected to keep filtering while the platform keeps supplying.

A focused community changes the terms of the encounter. It does not make every discussion thoughtful, and it does not guarantee that every recommendation will suit every reader. But when people arrive because they care about books, stories and reading, the conversation begins with a shared subject. That reduces one kind of noise: the need to establish why the subject matters at all.

Focus also makes it easier for a reader to be more than a consumer of recommendations. They can bring a question, a reaction or a half-formed opinion. They can describe a reading experience before they have turned it into a polished review. This is important because discovery is often social before it is certain. People do not always know what they think about a book until they have tried to explain it.

The risk of replacing one feed with another

There is a caution here for every platform built around discovery. A book-focused community can still become tiring if it recreates the same pressures: relentless posting, popularity contests, repeated talking points and the assumption that visibility equals value.

Focus is not automatically the same as quality. A smaller subject area can still produce noise if conversation is reduced to slogans or if readers feel they must perform expertise. The alternative to algorithm fatigue cannot simply be a new stream of content with more literary packaging.

The stronger alternative is a different culture of participation. Readers need permission to be curious without having an instant verdict. They need room for unpopular opinions, qualified recommendations and books that do not arrive with a ready-made identity. They also need to be able to leave a conversation with more possibilities than they had at the start, rather than with another set of instructions about what they ought to read.

Choosing with intention again

Algorithm fatigue may encourage a useful change in reading behaviour. Instead of asking only which book the internet has selected for us, readers can ask what kind of discovery they want today. Do they want something familiar or disruptive? A book that matches their usual taste or challenges it? A quick recommendation or a discussion with enough detail to make the choice feel informed?

Those questions do not make technology irrelevant. They make the reader’s intention more important than the feed’s momentum. Recommendations can remain part of the process, but they no longer have to be the whole process.

The future of book discovery may not depend on finding the perfect prediction. It may depend on restoring the human reasons that make recommendations meaningful: taste, memory, disagreement, enthusiasm and context. When readers can talk about what they are looking for, they are not merely being served more books. They are taking part in the culture that helps books travel from one person to another.