{"id":636,"date":"2025-11-06T18:08:09","date_gmt":"2025-11-06T10:08:09","guid":{"rendered":"https:\/\/www.chain258.com\/?p=636"},"modified":"2025-11-07T13:06:52","modified_gmt":"2025-11-07T05:06:52","slug":"study-warns-feeding-is-the-primary-killer-of-model-reasoning-ability","status":"publish","type":"post","link":"https:\/\/www.chain258.com\/index.php\/2025\/11\/06\/study-warns-feeding-is-the-primary-killer-of-model-reasoning-ability\/","title":{"rendered":"New Study Warns: Feeding AI Low-Quality Social Media Data Can Cause Lasting Cognitive Harm"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">\u200bA groundbreaking preprint study published on arXiv highlights a concerning phenomenon: when large language models (LLMs) are excessively trained on shallow, fast-moving, and often sensational content from social media, they can develop what researchers describe as &#8220;cognitive deterioration&#8221;\u2014a decline in both logical reasoning (IQ) and emotional judgment (EQ).<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The findings reveal that prolonged exposure to low-value online posts disrupts an AI\u2019s ability to think critically. Instead of processing information thoroughly, the models start taking shortcuts, omitting essential reasoning steps, or worse, dispensing with analysis entirely and generating outright wrong answers. The more &#8220;noise&#8221; in the training data, the more severe the degradation in performance.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">To assess the impact, researchers evaluated multiple open-source models, such as Meta\u2019s Llama 3 and Alibaba\u2019s Qwen, after training them on one million X (formerly Twitter) posts. The results showed that even initially well-balanced AI personalities began exhibiting stronger negative traits\u2014some even developed tendencies resembling certain unhealthy psychological patterns.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Efforts to reverse the damage through fine-tuning or introducing high-quality data yielded minimal improvement. The models retained their habit of rushed, superficial responses, proving that once poor reasoning patterns are learned, they\u2019re extremely difficult to unlearn. This underscores the importance of proactive data curation rather than relying on reactive fixes.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The study\u2019s central message is clear:\u00a0<strong>the quality of training data fundamentally shapes an AI\u2019s capabilities.<\/strong>\u200b Experts stress that future AI development must involve rigorous filtering mechanisms to prevent low-quality, noisy content from compromising model integrity.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">With major platforms like LinkedIn now exploring user-generated content for AI training, this research serves as a critical warning:\u00a0<strong>if we don\u2019t carefully filter the input, we risk creating not helpful digital assistants, but cognitively impaired AI systems.<\/strong>\u200b The solution isn\u2019t just better post-training adjustments\u2014it\u2019s ensuring a &#8220;clean diet&#8221; from the very beginning.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u200bA groundbreaking preprint stu&hellip;<\/p>\n","protected":false},"author":2,"featured_media":637,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[8,1],"tags":[160,24,26,159],"class_list":["post-636","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-deep-tech","category-uncategorized","tag-ability","tag-ai","tag-deep-tech","tag-low-quality-data"],"_links":{"self":[{"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/posts\/636","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/comments?post=636"}],"version-history":[{"count":3,"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/posts\/636\/revisions"}],"predecessor-version":[{"id":667,"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/posts\/636\/revisions\/667"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/media\/637"}],"wp:attachment":[{"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/media?parent=636"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/categories?post=636"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.chain258.com\/index.php\/wp-json\/wp\/v2\/tags?post=636"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}