TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

'Dangerous nonsense': AI-authored books about ADHD for sale on Amazon

56 点作者 n1b0m11 天前

6 条评论

elcritch11 天前
So a single company that claims to detect AI writing says these two linked books were AI written? The one book that was actually read doesn’t seem linked or the name actually given, so it’s hard to verify.<p>The article makes good click bait, but the substance seems lacking. The two actual linked books have decent reviews. Perhaps there were written by someone with ADHD who leveraged LLMs to help write a book that others also found helpful? Honestly I’m more worried about fake reviews, AI or human.<p>The content of the anonymous ADHD book that offended the person the article centers on? Well they apparently read seemingly horrible AI generated nonsense like people with ADHD are “four times more likely to die significantly earlier”. Well unfortunately that’s actually generally true. Maybe it’s AI written sentence but there’s plenty of research showing folks with untreated ADHD die earlier and are much more likely to die in accidents. I don’t care to quote research here but it’s readily found.<p>The other complaint about ADHD outbursts resulting in lasting scars? Well unfortunately that can also be true. Folks with ADHD are more prone to saying things they don’t mean or even if true said in a hurtful way.<p>One ADHD book I read years back pre-LLMs, was written by a counselor whose ADHD client snapped back something like “I’m glad your mother died” to their spouse without meaning too. The counselor said it took years for the couple to heal from that and never fully recovered. So yes, ADHD can result in injuries that leave lasting scars. Sometimes ADHD can really suck.<p>Overall the article seems to be “personal read unpleasant things about their ADHD diagnosis” and blames it on LLM tools with scant evidence. It’s about as annoying an article as posts on HN claiming another comment sounds AI written. Maybe it is, maybe not, but crap writing existed before LLMs. The article has less overall thought IMHO than an actual LLM might show.
fritzo11 天前
Is this just the more general problem of steadily decreasing cost to create &amp; distribute content leading to a longer tail of low-quality content, requiring stronger filtering mechanisms? I don&#x27;t see what&#x27;s specifically AI-related here, that wouldn&#x27;t also be a criticism of blogging, e-books, self-publishing, or even movable type.
评论 #43889288 未加载
评论 #43889555 未加载
TrackerFF11 天前
We very recently had a scandal here in Norway where a city council wanted to shut down a bunch of schools, and to justify this they had written a report where they cited research that backed up their rationale.<p>Problem was, someone had used some LLM to write the report, which in turn had hallucinated research...but used real authors as the foundation. Some observant journalist contacted the researchers, to inquire on their work - and the researchers were baffled, as they had never written those papers that had been cited in the report. Real world consequences of AI misuse.<p>But back to the topic of OP: I&#x27;ve seen some pretty sketchy science books, which tend to get spammed on various math FB groups. They look legit enough, but you soon discover that they&#x27;ve indeed been AI generated. Not anywhere near as dangerous as the books in the article, but still - imagine how many out there are reading AI generated slop, under the assumption that it is legit.
gedy11 天前
This was solved in past by publishing editors, bookstore owners&#x2F;purchasers, and even librarians.<p>It&#x27;s cool &quot;we have scaled&quot; to streamline the publishing direct to readers, but maybe profitable sellers like Amazon should like pay humans to review and filter out generated crap?
评论 #43889579 未加载
pengaru11 天前
The invasion of the bullshit generators is completely out of hand.<p>This level of &quot;AI&quot; is more like &quot;artificial well-formed stupidity&quot; and its volume of output is boundless. We&#x27;re totally fucked if it doesn&#x27;t become actually intelligent.<p>I&#x27;m already having to deal with clowns pasting useless chat bot drivel into GH comments and JIRA tickets at work instead using their brains and time to do actual work.<p>My colleague keeps joking about it being a good time to become a farmer, I&#x27;m starting to believe him.
评论 #43889679 未加载
评论 #43889549 未加载
pureagave11 天前
Just wait until the Guardian finds out about all the ADHD diagnostic centers, tutors, therapy mills and drugs. Some might say there is a giant money printing industry of nonsense around ADHD.
评论 #43889423 未加载
评论 #43889279 未加载