{"id":10229,"date":"2026-04-14T18:33:20","date_gmt":"2026-04-14T18:33:20","guid":{"rendered":"https:\/\/musictechohio.online\/site\/something-fundamentally-wrong-llms-communicate\/"},"modified":"2026-04-14T18:33:20","modified_gmt":"2026-04-14T18:33:20","slug":"something-fundamentally-wrong-llms-communicate","status":"publish","type":"post","link":"https:\/\/musictechohio.online\/site\/something-fundamentally-wrong-llms-communicate\/","title":{"rendered":"There\u2019s Something Fundamentally Wrong With LLMs"},"content":{"rendered":"
\n

Once you familiarize yourself with the characteristics of ChatGPT\u2019s writing, it becomes impossible to miss. The internet has been flooded with AI-generated text that often features distinctive language patterns<\/a>, from liberal use of em dashes and repetitive sentence structures to specific turns of phrase and tone.<\/p>\n

The trend has become so ubiquitous that experts are now warning that it could even influence the way we speak in real life.<\/p>\n

As historian Ada Palmer and cryptographer and author Bruce Schneier argue in an opinion piece for The Guardian<\/em><\/a>, it\u2019s a very real risk that could also entrench a fundamental flaw plaguing large language models today. While these models were trained on vast quantities of written text, social media posts, movies, TV shows, and other recordings, this data often comes up short in \u201cunscripted conversations we have face-to-face or voice-to-voice\u201d \u2014 which represents the \u201cvast majority of speech, and a vital component of human culture.\u201d<\/p>\n

It\u2019s a massive blind spot that could result in humans eventually adopting linguistic patterns of these models, among other far-reaching consequences.<\/p>\n

\u201cThis will affect not just how we communicate with one another,\u201d Palmer and Schneier write, \u201cbut also how we think about ourselves and what goes on around us.\u201d<\/p>\n

\u201cOur sense of the world may become distorted in ways we have barely begun to comprehend,\u201d they concluded.<\/p>\n

Research has already shown<\/a> that AI-generated language relies on shorter-than-average sentences, while using a narrower vocabulary than human speech. It also sacrifices what makes human-written text human, including what Palmer and Schneier term \u201cmeanders, interruptions and leaps of logic that communicate emotion.\u201d<\/p>\n

Worse yet, AI models developed after the advent of ChatGPT run the risk of being trained on output that was itself generated by an AI, a dangerous feedback loop<\/a> that could further entrench these machine-inspired patterns.<\/p>\n

Beyond linguistic choices, AI models have long been shown to be highly agreeable or \u201csycophantic<\/a>\u201d towards the user, often indulging in their potentially flawed or downright dangerous<\/a> lines of thinking or beliefs. It\u2019s a tendency that can \u201creinforce bias and even worsen psychosis,\u201d Palmer and Schneier argue.<\/p>\n

For impressionable minds, the consequences could be far reaching. Educators are warning that students are losing their ability to think for themselves<\/a>, instead choosing to consult AI when prompted with a question they can\u2019t answer. University students are worried their peers are starting to all sound the same<\/a>, relying on the same machine-generated output. Meanwhile, experts are worried the widespread use of AI products in the workplace could be causing users\u2019 cognitive faculties and critical thinking skills to deteriorate<\/a>.<\/p>\n

Finding a solution in the long term to have these AI models better reflect when we\u2019re \u201cat our most authentically human\u201d could prove difficult. However, that shouldn\u2019t stop us from looking for one.<\/p>\n

\u201cWe don\u2019t pretend to know what the best solutions might be,\u201d Palmer and Schneier concluded in their piece. \u201cBut one has to imagine if there\u2019s ingenuity to develop AI models, then surely there\u2019s ingenuity to come up with a way to train them on informal human speech instead of us only at our most stylized, veiled, and sometimes worst.\u201d<\/p>\n

More on AI:<\/strong> Large Language Models Will Never Be Intelligent, Expert Says<\/em><\/a><\/p>\n

The post There\u2019s Something Fundamentally Wrong With LLMs<\/a> appeared first on Futurism<\/a>.<\/p>\n<\/div>\n

<\/div>","protected":false},"excerpt":{"rendered":"

Once you familiarize yourself with the characteristics of ChatGPT\u2019s writing, it becomes impossible to miss. The internet has been flooded with AI-generated text that often features distinctive language patterns, from…<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[177,3841],"tags":[],"class_list":["post-10229","post","type-post","status-publish","format-standard","hentry","category-artificial-intelligence","category-ethics"],"_links":{"self":[{"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/posts\/10229","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/comments?post=10229"}],"version-history":[{"count":0,"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/posts\/10229\/revisions"}],"wp:attachment":[{"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/media?parent=10229"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/categories?post=10229"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/musictechohio.online\/site\/wp-json\/wp\/v2\/tags?post=10229"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}