Google Books is reported to be indexing bad books generated by AI

Google Books, a vital resource for scholars, has initiated the indexing of substandard books. This development may have effects on how Google Books provides information to its language monitoring tool, Ngram.

404Media publishes Google Books containing several volumes that could have been written by artificial intelligence. The article conducted a Google Books search using the phrase “as of my most recent knowledge update,” which is frequently employed by chatbots such as ChatGPT. A search on Google Books for particular phrases or terms will typically return written works that contain those terms or sentences.

It was discovered that the majority of the books on the first few pages of the search results were about artificial intelligence; however, there were a few that did not discuss the technology and appeared to have been written by a computer.

404Media reported that the books it discovered, including Bears, Bulls, and Wolves: Stock Trading for the Twenty-Year-Old by Tristin McIver, appeared to have combed through Wikipedia for information on financial events and did contain the phrase “as of my most recent knowledge update.” Additional volumes about subjects such as Twitter continued to include content from 2021, the most recent time certain AI models received training data.

The majority of the data supporting Ngram, a research instrument that monitors the evolution of language, is sourced from Google Books. Ngram uses data extracted from written materials to illustrate the evolution of language usage.

Written works from the 1500s onward were scanned and indexed by Google Books. The most recent update to the data cited by Ngram was in 2019. Numerous linguists and other academicians utilize Ngram for research gathering, despite its imperfections.

Recent works on Google Books do not appear in Ngram results, Google informed 404Media; however, future data updates may include these works.