Amazon’s AI-written mushroom foraging books could be ‘life or death’

Amateur mushroom enthusiasts have received a cautionary warning to steer clear of foraging guidebooks listed on Amazon that bear the marks of having been authored by AI chatbots. Foraging experts have told How To Be Books that “malicious actors” are exploiting the ease of AI-generated books at the expense of unwitting members of the public, and that toxic species could be recommended as a result of lack of due diligence.

However, it’s not the first time Amazon has been questioned for its AI works. Amazon, known as a sprawling marketplace for various literary offerings, has inadvertently become a platform for these types of publications masquerading as human-written works, with genres like science and education often being compromised by these deceptive entries.

Examples of potential AI written mushroom foraging books

In the realm of wild mushroom foraging, an emerging concern has emerged as a series of books cropped up on Amazon’s virtual shelves. Possible AI-generated books like “Mushroom Farming for Beginners 2023” and “New Field Guide for Edible Medicinal Mushroom Growing, Foraging & Gourmet Farm Cultivation” have appeared, with the latter title subsequently being pulled from the platform.

Upon even cursory scrutiny, it becomes apparent that AI technology played a significant role in crafting these texts. Notably, in “Mushroom Farming for Beginners,” a simple query posed to the ChatGPT AI model— “what is mushroom farming?”—elicited remarkably similar results, underscoring the automated nature of the content creation process.

Original text:

ChatGPT query

AI detection

As a result, we tested the veracity of these writings using Copyleaks, an AI-powered detection tool, which revealed staggering probabilities of AI involvement. The tool suggested a likelihood of 93.4% that AI played a role in generating the paragraph in question. The content from the book “New Field Guide for Edible Medicinal Mushroom Growing, Foraging & Gourmet Farm Cultivation” exhibited a parallel pattern, with Copyleaks attributing an 83.6% probability to AI composition.

Exploring the AI-written passages further, it becomes evident that the language and structure bear striking resemblances. For instance, the text elucidates the concept of mushroom farming: “Mushroom farming, also known as mushroom cultivation or mushroom growing, is the process of growing edible mushrooms for commercial or personal use. It involves the controlled cultivation of various species of mushrooms in a controlled environment, such as a greenhouse or indoor facility.”

Original text:

ChatGPT query

AI detection

The issue compounds when delving into another taken-down work, the “New Field Guide for Edible Medicinal Mushroom Growing, Foraging & Gourmet Farm Cultivation.” The text also expounds upon medicinal mushrooms’ traditional applications in Chinese medicine and the extraction of active metabolites for supplement use.

AI-generated mushroom books raise concerns among mycology and foraging experts

The guidebooks have obviously sparked concerns within the mycology community, as enthusiasts and experts alike voice worries over the accuracy and safety of these deceptive publications. The exposure of this phenomenon began with the 404 Media site, which shed light on the infiltration of Amazon’s virtual shelves by AI-authored books masquerading as works of human origin.

One organisation at the forefront of addressing these concerns is the New York Mycological Society. Notably, the society turned to X, formerly known as Twitter, to convey their concerns and issue a warning to potential readers: “@Amazon and other retail outlets have been inundated with AI foraging and identification books. Please only buy books of known authors and foragers, it can literally mean life or death.”

Elan Trybuch, secretary of the New York Mycological Society, provided insights into the unfolding situation. Speaking to How To Be Books, Trybuch said he was first made aware of it through Reddit, and working in the software field, was already conscious of Al’s ability to “manipulate facts or even blatantly lie to feign knowledge.” As a result, he shared the warning with the organisation’s followers.

Asked why there had been an influx, Trybuch highlighted a critical concern that goes beyond the realm of profit-driven motives.

Sigrid Jakob, who is president of the society, also shared her thoughts, saying that there has been growing interest in foraging in general since the pandemic.

Read: Authors’ pirated books used to train Generative AI

However, the repercussions extend beyond financial gain, delving into potential hazards. Trybuch said AI is unable to “distinguish between a poisonous mushroom, like the Death cap (Amanita bisporigera), and a non-poisonous variety, like the Wood Mushroom (Agaricus silvicola).” He claimed the society’s members apparently encountered individuals in Harriman State Park who, prior to AI books, were foraging for the toxic Amanita bisporigera.

‘Fortunately, we intervened and educated them about the differences. During our excursions, we employ methods like using KOH (Potassium Hydroxide) to detect cap color changes.” Hence, Jakob reiterated that the consequences of misinformation or poorly-worded guidebooks “might mislead guidebook users into deeming a toxic species as safe to consume.”

Read: Amazon halts AI-generated books impersonating author Jane Friedman

Trybuch presented a proactive approach to combat the surge of AI-generated works by holding identification sessions, which can be held online or in-person. They also issue a quarterly newsletter features member-submitted book reviews. “Members collaborate to share knowledge and resources. Through these avenues, the spread of AI-generated books can be curtailed,” he added. In addition, Jakob suggested consulting experts and joining local mushroom clubs before embarking on foraging.

While these AI-authored books raise pressing concerns within the mycology community, Amazon’s involvement in addressing the issue has also come under scrutiny. Amazon spokesperson Ashley Vanicek told us: “Amazon is constantly evaluating emerging technologies and is committed to providing the best possible shopping, reading, and publishing experience for our authors and customers. All publishers in the store must adhere to our content guidelines, regardless of how the content was created.

“We invest significant time and resources to ensure our guidelines are followed, and remove books that do not adhere to these guidelines. We’re committed to providing a safe shopping and reading experience for our customers and we take matters like this seriously.”

Consequently as the dialogue continues, it remains to be seen how online marketplaces and consumers navigate the intersection of AI-generated content and accurate, trustworthy information.