Among the titles offered: Navigating ADHD in Men, Men with Adult ADHD: Highly Effective Techniques for Mastering Focus or Diet & Fitness for Adult ADHD. All analyzed by Originality.ai, an American detection tool. Verdict: 100 % score on the probability scale that the text is generated by an AI.
The phenomenon worries specialists. For Michael Cook, computer researcher at King’s College in London, these contents can be dangerous: “Generative AI may recommend the ingestion of toxic substances or ignore essential medical instructions. He deems worrying their growing presence on platforms, especially when it comes to health subjects.
False experts, real dangers
According to him, the AI have neither reliable memory nor the necessary critical discernment: “They are formed as much on medical works as on conspiracy theories or pseudo-science“Reports the Guardian.
Amazon, for its part, takes advantage of each sale. “”The economic model encourages these practices, without the producers of these content being worried“, Continues Cook.
Shannon Vallor, professor at the University of Edinburgh, abounds in this direction. For her, Amazon has a moral responsibility: “It is not a question of making a bookseller responsible for all the content sold, but he cannot ignore the drifts he facilitates. »
Toxic advice sold as an expert guides
Traditional publishing safeguards-rereading, reputation, verification of the authors-collapse under the weight of AI. And the legislative vacuum around these practices promotes a race for performance, even if it means misinfreming.
-Richard Wordsworth has had the bitter experience. Recently diagnosed with adult ADHD, he buys a book recommended by his father. From the first pages, he felt the scam: quote from Jordan Petersen, incoherent anecdotes, historical errors, and anxiety -provoking affirmations.
A passage Participation particularly: it is said that those around you do not forgive the “emotional damage” caused by impulsive anger. He then discovers that the author probably does not exist, his portrait likely generated artificially.
« My father, however very cultivated, was fooled. If he can believe it, everyone can. Vulnerable people find themselves reading toxic absurdities written by crooks while Amazon cashes“, He deplores.
Amazon nevertheless ensures to monitor his content: “We have strict guidelines, detection tools, and we remove non -compliant books. Our processes will evolve with the landscape of publishing. »
Photo credits: Shutterstock
By Clément Solym
Contact : [email protected]
Related news :