Use AI to help with reading
In my free time this past year, I rediscovered the joy of reading, especially novels.
Reading novels does take time and a mindset, because it is slower to get into a flow state. If you can’t enter the state for a long time, it’s easy to stop reading; but once you get into it, it is more immersive and memorable than playing games (interactive form) or watching movies and TV series (multimedia form). You can think over and over about the beautiful parts and appreciate the emotion, and it’s easier to stop and think about the empty spaces the author left in the plot. Reading speed is completely under your control, whether fast or slow. Considering their production costs, novels are much more diverse than other media and offer a wider choice.
I recently tried using AI to improve my reading experience. The first thing I discovered is that AI is great for recommending books. I mainly use Gemini, the free version is enough. I can first list some books that I really like and let him recommend more to me. Among the selected candidates, learn the features of this book through dialogue. To avoid reading similar books over and over again, I’ll also ask the AI to recommend some genres I’ve never tried before. Of course, the novel itself is still created by people, and it is more efficient to recommend the author than the book itself.
In the past two months, I wanted to read some space opera-type novels, but I’ve basically read all the old classics, so I turned to new works from the past ten years. Another reason is that most science fiction novels are time sensitive. In recent years, science and technology in human reality have developed rapidly, and a writer’s fantasy can easily lose touch with reality as time goes by.
But I quickly discovered that the biggest problem in reading new science fiction novels was that the translation speed of the Chinese version couldn’t keep up. 90% of books recommended by AI do not have Chinese translations. Even if we expand the time horizon a little, long novels written ten years ago are often only the beginning of them translated. This is easy to understand: if sales of the first book published are less than satisfactory, it is conceivable that subsequent books will be even less satisfactory. This is absolutely economically rational behavior, but quite inconvenient for his fans.
As a side note, the board game field is somewhat similar. Table games are often designed by one person and are not intended for the general public. Even if the designer has the series in mind, if the previous games don’t sell well enough, it will be difficult to release an expansion pack. My favorite tabletop game designer Thomas Lehmann explains why Res Arcana’s third expansion, Res Arcana Duo, was released as a (seemingly simplified) standalone game instead of continuing the expansion pack format: There had to be a way to expand the player base of the series, otherwise expansion pack sales would dwindle. As a Chinese user, I’m still a little sad that Res Arcana Duo hasn’t been able to release a Chinese version. I hope that the new work Dark Pact (2026) this year can be released in Chinese. After reading the introduction, I was quite interested in this pure card building game.
I’m a big fan of the Old Man’s War series, and I really like John Scalzi. The book is not at all tiring to read, and the cool programmer-style humor that is seen everywhere in this book is quite to my liking. I bought the first book in the Mutual Empire series “Collapse of Empire” on JD.com a while ago. It was five o’clock in the afternoon when I received the book on a warm day, and I closed the book at ten o’clock in the evening. Apart from the usual dinner, I spent the rest of the time reading. Another motivation for reading this book was that I wanted to see more different settings surrounding space travel (to provide inspiration for my space travel games). After reading this book, apart from being very satisfied with the science fiction setting in the book, I am also looking forward to further story developments.
Unfortunately the last two books of this trilogy were never translated into Chinese.
I feel like my English reading level has improved a lot in the past few years, so why not try reading English directly? I tried, but was far from enjoying reading. Reading novels requires a smooth experience. If you can’t get into the flow smoothly, reading becomes a chore. There is a big difference between literary works and technical articles. Being able to read technical English fluently doesn’t mean you can read novels fluently. I think more reading practice is needed, and studying must be difficult, and this is not what I want right now.
Two days later, I tried another method, which was another important help I discovered AI can provide: reading translations.
In my opinion, the biggest difference between technical articles and literary works is that technical articles aim to express knowledge concisely and accurately, while technical articles need to convey emotions in addition to describing the storyline conceived by the author. Understanding a novel requires understanding the characters and story in the novel; just like translating a technical book, you have to understand the technical principles. This is why machine translation cannot be accurate. Large language models are supposed to improve translation, but what I initially tried was Google Translate directly and a local Translategemma model installed in local ollama, for literal translations of novels.
After using different methods and experiencing several chapters, I found that it was best for me to have the machine translate them in full, without any processing for Chinese context, and display the sections in Chinese and English for me to read. I mainly read in Chinese. Although the language felt a bit lame, because I knew that the information was originally in English, and I had sufficient knowledge of English grammar, my brain was able to adapt quickly and automatically switch to appropriate Chinese comprehension during the reading process. Since this is a mechanical literal translation, no information will be lost. When you find a sentence difficult to understand, immediately jump to the original English text and it will usually become clearer. When reading beautiful dialogue, it is often more moving to reflect on the original English sentences.
Some sentences are quite difficult to understand. Currently, you can open a Gemini conversation, provide enough context, then paste the original text. Gemini can explain it very clearly. After all, this is a novel written 10 years ago, and I guess the original text of the novel itself (or even the Chinese translation of the first volume) is training material for the grand model. For example, I learned something this time: in an English context, the emperor would call himself Kami/Us, not I/Me, which is used to refer to the dual identity of the individual and the imperial power behind him. This is very similar to the Chinese context where the emperor calls himself “I”. When I read the English literal translation for the first time, I was confused by “our” output from the translator, but I learned it after discussing it with the AI. In the Chinese translation of the first volume, the translator correctly chose “朕” to translate We. Google Translate is clearly not a literal translation, but Gemini can choose this translation method according to the context. I strongly suspect that it is influenced by the training corpus (trained with Chinese subtitles).
I finished reading the second book “Consuming Fire” this way, and it took about 2-3 times longer than the first. The decrease in reading speed is noticeable, but acceptable. I think with a little training, this could be improved to the point where the reading flow wouldn’t be affected at all. Then I read the third book, “The Last Emperox”, and it took just as long as the first book. However, I think it’s not because I quickly adapted to this new reading method, but because the story structure of the three books in this series is actually similar. After reading the book, following the rhythm of the book became easier. The process of reading a novel is like gradually building the world created by the author in your mind, then filling in the details little by little. The hardest part lies ahead, and the rest is a smooth process.
Even though the plot is somewhat similar, I still really like this trilogy.
I have finished the last three volumes of the novel “The Expanse” in the last two days, so there is no need to wait for the American TV series 🙂
Berita Terkini
Berita Terbaru
Daftar Terbaru
News
Jasa Impor China
Berita Terbaru
Flash News
RuangJP
Pemilu
Berita Terkini
Prediksi Bola
Technology
Otomotif
Berita Terbaru
Teknologi
Berita terkini
Berita Pemilu
Berita Teknologi
Hiburan
master Slote
Berita Terkini
Pendidikan
Resep
Jasa Backlink
Slot gacor terpercaya
Anime Batch
