We need to talk about AI

This post may include affiliate links, which means we have made a small commission on any sales. This committee can help the Feminist Book Club pay our contributors, so thank you for your support of the small independent media!

or one Actual Summer reading list…

By now, you’ve heard of the AI-generated summer reading list, which is published in Chicago (print!) newspapers and on multiple news sites. You may also hear that two-thirds of this reading list are completely fictional.

Apart from this, it’s not true. Every author on the Summer Reading List is a truly bestselling author. Some of these books are also real. look: Hurricane seasonattributed to Brit Bennett (author Half of disappearing), but actually written by Fernanda Melchor (and former FBC draft picks). The attached introduction is completely fictional, at least related to the book and the two authors. The longest day It’s a real book, but it’s Cornelius Ryan and not Rumaan Alam’s book.

Strangely, all “previous works” are attributed to the authors in the summary yes Really, correctly attributed to 8 times! It is also correct to mention that the author’s nationality/expat and the genre they tend to write about. So, how does AI make this list so wrong? How to catch it in the research, writing and editing process?

How did this happen?

The book list that causes all of these Hubbub is a joint article that has been licensed from content providers and operated by multiple publications. Joint organizations are not new and are commonplace in all types of news media, from printing paper to radio to television.

Explained that the problematic summer reading list appeared, and unfortunately, the papers announced the loss of 20% of the writers due to financial difficulties of the parent company. However, despite the reduced payment capacity, there is no need for content.

This is where the group content comes in. Since it is universal enough to be published in multiple places without much hassle, it can be easily used to fill content gaps in publications. However, the fact remains that these contents still require research and human hands to produce.

Or we think.

Unusually, this particular summer reading list seems to be completely Production by some form of AI large language model (such as Chatgpt). In fact, Marco Buscaglia, the author in charge of the work, took some responsibility to show that he used AI for research purposes, although he claimed that the article was “partially” by AI.

Good book content requires more than button press

As a content creator, it’s hard for me to believe it. I like to think that I won’t post or even submit anything in the first two thirds is absolutely fictional nonsense (although some of my podcasts of course!). But even though I did, despite being part of a relatively small team, I still had serious mistakes during the editing process.

Furthermore, we tend to actually read books recommended to our audience, or at least we studied them and their authors.

The crux of the problem lies in: It takes more time and effort to produce content than the time and energy that ordinary consumers realize. For example, I’m The Stone Witch of Florence I was asked to read a 362-page novel and take notes at the same time. Then I drafted this article, which could take several hours. I also made the pictures that appeared and formatted everything in the draft. Finally, I submitted it to our editorial team.

However, this is not the end of the work. Our editor read the comments and contacted me through edit requests and questions, a process that could span multiple days before eventually scheduling the post. This prompted a third person to prepare additional content for our social media accounts and online communities and announced new blog posts.

As you can see, it will take several weeks to spread the 4-minute blog posts to the world. So imagine the recommended potential job List books. Now imagine that free content creation and journalism are your main income. You can see the temptation to cut off the corner.

The impact of artificial intelligence continues to grow

Today, many people are eager to outsource their ideas to large language model AI, allowing Chatgpt to do their homework, write papers, and even write shopping lists for them. However, part of the problem is that from its core, large language models are really just pattern recognition software. They are not smart and do not know how to distinguish the correct information from the incorrect information provided through training packets. Training data sets are arbitrarily accumulated from the Internet, including Reddit and other places. There are almost no fortresses except cats (note: I love Reddit and cats; don’t come for me in the comments!). The point is that AI will only fill in the blanks with anything that looks like it can.

Unfortunately, some people treat Chatppt like Google, and they have no fact-checked results. This is how we get 10 books with complete makeup. (Or how we have trouble with theft.)

In addition to the morality brought about by generating AI, which may be a question of morality and certain accuracy, it also carries a more serious problem: excessive consumption of the environment. Generating AI relies on a large number of data centers to process the required large amount of data sets and meet user requests. Such data processing requires a large number of servers, which generate a lot of heat. These systems need to be cooled to remain operational. Do you know how the SNES system doesn’t have a fan and will overheat before completing the Toad fight? Well data centers are at the same risk, and they use water from local resources to keep them cool, which creates huge water availability issues for those who already live there. Then there is power consumption, which is so much of a concern that Microsoft has reached a deal to restart Three Mile Island (a previously closed power plant) and buy all the electricity generated over the next 20 years.

So we are here, depriving people of drinking water for selfish reasons and reopening nuclear power plants, so people can all replace Google with Chatgpt, quit their therapists from work, steal art because AI needs to steal, or even use it to write books instead of using integrity and imagination.

Anyway, this is a summer reading list of real and actual books I’ve actually read.

What should you read this summer

The baker and magic fellowship J. Penner: A slow-burning romantic film… Great British Baking Meet high fantasy. For fans of Travis Baldree, Heather Fawcett and Coyote JM Edwards.

In front of the big bird Hiromi Kawakami: Cli-Fi encountered a weird FIC. For Becky Chambers fans and Horror, Netflix series Love, Death and Robots Or Sun-punk-style stories. If one of your New Year resolutions involves wanting to read more translated works, please select this. See the podcast discussion on the International Booker Award shortlist to hear me talk about this.

Own Alba Diaz Isabel Cañas: horror/romance, property, yearning. Historical novel. Dark, strong man prototype. For fans Mexican gothic and Northern Vampire. If you like Bochica Written by Carolina Florez-Cerchiaro, you’ll love it.

All other mothers hate me Sarah Harman (to be adapted into TV): The irony of mom’s life. The mystery of the murder of female anti-heroes. Set in a part of London’s wealthy and with a wealthy role. For fans Violators (2021 TV program) and crow Danielle Page and Kass Morgan. (Please check out the podcast episode, where Mariquita and I discussed the book and other books featuring “No Forest” women.)

The last few days of Midnight Walker Sarah Tomlinson’s author: A fictional memoir in which the ghostwriter produced “behind the scenes” for the fictional rock band Muse of The Fival Rock. Despicable, charming and dramatic. For fans Daisy Jones and Six Literally, everything Joan Jett has done.

The list is based primarily on resonance, you don’t get this humanity from anyone, I don’t care about the level of intelligence it should have.