Show HN: I wanted better book recommendations – so I made Lorekeep | Hacker News Reader

Show HN: I wanted better book recommendations – so I made Lorekeep

lorekeep.io

19 pointsdelta23411 comments

I often have the issue that AI recommendation algorithms give me more of the same. I already know what I like.

I've found that when I read books in genres that I'd never thought of reading (for instance, biography of a writer in 1800 France) that were actually much more enriching and valuable additions to my bookcase.

I already know where to find the books that I know I will like for the most part.

Some sites seem to really struggle with recommendations - after looking at 20+ washing machines, over a period of weeks, I bought one recently.

I'm a normal adult and now I will probably not even consider buying another washing machine for the next ten years, and yet my recommendations? More. Washing. Machines.

Many recommendation-sites are exactly like that - I read books 1-4 of a particular author, so of course I want the same genre, along with every other book by that same author.. Terrible.

The concept is great. Some constructive feedback on the execution:

I expected to be able to provide a lot more than just five books. The splash page says "Our AI analyzes your reading DNA to recommend books you'll absolutely love—guaranteed to match your unique taste." That makes it sound like it's going to dive deep into my reading history and really think about my taste. Five books is absolutely not "my reading DNA."

There are lots of ways to get a richer picture of the books a person loves. You could connect to Goodreads or Storygraph, or scrape their social media for books you've discussed, or let users upload a .csv exported from other sources like LibraryThing, their Amazon wishlist, or their own local lists they might keep on Obsidian, Notion, or wherever. My public library keeps my reading history automatically - that would be another good data source for my "reading DNA."

Right now, it's just an AI recommendation based on five books. I can do that with any LLM from ChatGPT to Copilot to Gemini. The recommendation I got was very basic, just obviously similar books from the same authors I entered or ones that are closely related.

People's tastes are complex. Even if you allow much larger data sets to create a person's reading DNA, that alone won't necessarily recommend books that are right for them. For example, I love PG Wodehouse, but I have no interest at all in Evelyn Waugh, James Thurber, or G. K. Chesterton. A great recommendation engine would ask me why I love a book, and try to tease out the reasons behind my reading list in order to recommend books that will be more accurate and unexpected than I could get from a simple ChatGPT query or my Goodreads profile.

A site like this needs to do a lot to stand out. It's an excellent concept, I hope you develop it into something special.

One huge challenge is in identifying and measuring the axes of appreciation.

Asking "What do you like about X?" is a tough way to extract good data. People usually cannot explain why they like things. And it's legitimately difficult to know sometimes.

Also, tastes are often context-driven/sensitive. A book that I loved when I read it last summer in Barcelona, or on the 16-hour flight to Auckland ... does not necessarily map to what I would enjoy reading right now. Or that I should pack for my trip next week.

I've tried to suss this out in music. Songs are theoretically more approachable than books/films/etc: Bite-sized consumption quanta, a fairly robust (but large) genre taxonomy, one basic grounding theory (not really, but a reasonable approximation for the culture within which I exist). Then you can split out by instrumentation, style, arrangement, tempo, etc and get some well-defined groups.

This doesn't work. It's over-analytical, and under-representative of human taste spectra.

The "best" engines use high-resemblance cohorts, but no one actually likes them -- they give lame obvious suggestions, and are terrible at surfacing surprises. They're OK at "good enough, sometimes" in the same way that turning on a TV for the 6pm news and sitting there on the same channel until Letterman signs off was "good enough" (i.e. horrifically bad!) back when serial TV was a thing.

There remains something ineffable about taste -- "It don't mean a thing, if it ain't got that swing". (Ironically, "swing" is now probably measurable! But the point remains for other as-yet-undefined axes.)

That's a great point. For example, my wife probably knows me better than anyone in the world. She's very good at seeing a book and knowing I'm probably going to like it. That includes your example of "this is a good book for a trip" vs "this is a good book to read at home, at night." But even she gets it wrong about 20% of the time.

In order to be able to really recommend something as multi-faceted as a book, movie, or song, you have to know a person on pretty much every level. I suppose seeing a person's entire social graph, search history, LLM history, media consumption history, and browser history might get you close, but it's still a Hard Problem™.

I really like the idea, even the ISBN search works. However, I noticed that when I search for Nietzsche, for example, not all of his works are listed.

I asked myself: Which book databases does the application use? And what are/were the challenges? From my own experience (e.g. application that digitizes the offline book stock) I had a hard time finding open book databases that are well maintained and complete. IIRC at the time google and open library was the best solution besides some commercial ones.

In general, I have to say: I'm a bit lazy when it comes to choosing books. So when I'm not sure what to read, I just email my bookseller and ask for recommendations. The nice thing is, I don't have to name an author or title. They know me. If I decide to order a book, I just send them another email (maybe I can use your app to send them the ISBN and correct Title and Author). They then let me know when the book has arrived in the store and I pick it up. Sometimes I browse through the books in the store, or ask for recommendations.

Another way I get book recommendations is through friends. Over lunch, we talk about what we've just listen to or read.

And I have to admit that I listen to 2-3 audiobooks a month, the app gives me recommendations.

I've also thought about exporting my data from the audiobook app and letting AI give me recommendations. But at the moment I've bought more books than I bought time to read them.

Someday, someone will create a recommendation system that is more than mere simple pattern matching.

I enjoyed The King Must Die because it is an excellent novel about personal power, the nature of men and violence, and the duties of a leader. It was written by a woman who had a lot of wisdom about the male psyche.

But all I got were recommendations of other novels based in Greek mythology settings.

How do you find these recommendations? There's no explanation on the website.

It also says also reading data is private, but where's the privacy policy?

This seems, at least at the surface, to be strictly worse than going to chatgpt

Tried to use it, getting:

    Application error: a client-side exception has occurred while loading www.lorekeep.io (see the browser console for more information).

Looks like you need to handle some null cases from your API response. Console:

   Uncaught TypeError: Cannot read properties of null (reading 'description')
       at 870-062d78834b42ed0a.js:1:924
       at Array.map (<anonymous>)
       at o (870-062d78834b42ed0a.js:1:904)
       at lS (4bd1b696-75688abf446dcbb2.js:1:39319)
       at ot (4bd1b696-75688abf446dcbb2.js:1:64567)
       at ov (4bd1b696-75688abf446dcbb2.js:1:75632)
       at ic (4bd1b696-75688abf446dcbb2.js:1:112354)
       at 4bd1b696-75688abf446dcbb2.js:1:112199
       at is (4bd1b696-75688abf446dcbb2.js:1:112207)
       at u9 (4bd1b696-75688abf446dcbb2.js:1:109418)
       at iH (4bd1b696-75688abf446dcbb2.js:1:130644)
       at MessagePort.w (684-e6424b1559edc7ed.js:1:26075)
   (anonymous) @ 870-062d78834b42ed0a.js:1
   o @ 870-062d78834b42ed0a.js:1
   lS @ 4bd1b696-75688abf446dcbb2.js:1
   ot @ 4bd1b696-75688abf446dcbb2.js:1
   ov @ 4bd1b696-75688abf446dcbb2.js:1
   ic @ 4bd1b696-75688abf446dcbb2.js:1
   (anonymous) @ 4bd1b696-75688abf446dcbb2.js:1
   is @ 4bd1b696-75688abf446dcbb2.js:1
   u9 @ 4bd1b696-75688abf446dcbb2.js:1
   iH @ 4bd1b696-75688abf446dcbb2.js:1
   w @ 684-e6424b1559edc7ed.js:1Understand this error

API response:

    {
        "books": [
            {
                "id": "KaJBDAAAQBAJ",
                "title": "The Thief and the Dogs",
                "author": "Naguib Mahfouz",
                "description": "Naguib Mahfouz's haunting novella of post-revolutionary Egypt combines a vivid pychological portrait of an anguished man with the suspense and rapid pace of a detective story. After four years in prison, the skilled young thief Said Mahran emerges bent on revenge. He finds a world that has changed in more ways than one. Egypt has undergone a revolution and, on a more personal level, his beloved wife and his trusted henchman, who conspired to betray him to the police, are now married to each other and are keeping his six-year-old daughter from him. But in the most bitter betrayal, his mentor, Rauf Ilwan, once a firebrand revolutionary who convinced Said that stealing from the rich in a unjust society is an act of justice, is now himself a rich man, a respected newspaper editor who wants nothing to do with the disgraced Said. As Said's wild attempts to achieve his idea of justice badly misfire, he becomes a hunted man so driven by hatred that he can only recognize too late his last chance at redemption.",
                "thumbnail": "https://books.google.com/books/content?id=KaJBDAAAQBAJ&printsec=frontcover&img=1&edge=curl&source=gbs_api",
                "isbn": "9781101974650",
                "publishedDate": "2016-06-15",
                "categories": [
                    "Fiction"
                ],
                "pageCount": 162,
                "publisher": "Anchor"
            },
            null,
            {
                "id": "SKBBDAAAQBAJ",
                "title": "The Cairo Trilogy",
                "author": "Naguib Mahfouz",
                "description": "Nobel Prize-winner Naguib Mahfouz’s magnificent epic trilogy of colonial Egypt—Palace Walk, Palace of Desire, and Sugar Street—together for the first time in one beautiful hardcover volume. The masterwork of the Nobel Prize-winning author, the three novels of The Cairo Trilogy trace three generations of the family of tyrannical patriarch Al-Sayyid Ahmad Abd al-Jawad, who rules his household with a strict hand while living a secret life of self-indulgence. Palace Walk introduces us to his gentle, oppressed wife, Amina, his cloistered daughters, Aisha and Khadija, and his three sons–the tragic and idealistic Fahmy, the dissolute hedonist Yasin, and the soul-searching intellectual Kamal. Al-Sayyid Ahmad’s rebellious children struggle to move beyond his domination in Palace of Desire, as the world around them opens to the currents of modernity and political and domestic turmoil brought by the 1920s. Sugar Street brings Mahfouz’s vivid tapestry of an evolving Egypt to a dramatic climax as the aging patriarch sees one grandson become a Communist, one a Muslim fundamentalist, and one the lover of a powerful politician. Throughout the trilogy, the family’s trials mirror those of their turbulent country during the years spanning the two World Wars, as change comes to a society that has resisted it for centuries. Filled with compelling drama, earthy humor, and remarkable insight, “The Cairo Trilogy extends our knowledge of life; it also confirms it” (The Boston Globe). Everyman's Library pursues the highest production standards, printing on acid-free cream-colored paper, with full-cloth cases with two-color foil stamping, decorative endpapers, silk ribbon markers, European-style half-round spines, and a full-color illustrated jacket. Contemporary Classics include an introduction, a select bibliography, and a chronology of the author's life and times.",
                "thumbnail": "https://books.google.com/books/content?id=SKBBDAAAQBAJ&printsec=frontcover&img=1&edge=curl&source=gbs_api",
                "isbn": "9780525432029",
                "publishedDate": "2016-06-15",
                "categories": [
                    "Fiction"
                ],
                "pageCount": 1369,
                "publisher": "Everyman's Library"
            }
        ]
    }

I wanted better book recommendations - so I made AI slop generator... do we really need it.

It doesn't give any reasoning why it would be a good recommendation. Just basically random book from the same genre.

Reply on news.ycombinator.com