Elon Musk recently stated that his AI company xAI will retrain its AI model Grok on a brand new knowledge base free of "garbage" and "uncorrected data"—starting with rewriting history.
In a post released on the X platform on June 21, Musk pointed out that the upcoming Grok 3.5 model will possess "advanced reasoning capabilities" and aims to use it to "rewrite the entire corpus of human knowledge, supplement missing information, and eliminate erroneous content."
He further explained that the model will subsequently be retrained based on this new knowledge set, while claiming that "there is an excess of useless information in any base model trained on uncorrected data."
Musk's latest confrontation with the "woke" faction
Musk has long argued that competing AI models like ChatGPT from OpenAI (which he co-founded) are biased and deliberately ignore politically incorrect information.
For years, Musk has been committed to creating products free from what he considers harmful political correctness, striving to shape Grok into what he calls an "anti-woke" tool.
After taking over Twitter in 2022, he significantly relaxed the platform's content moderation and misinformation oversight standards, leading to an influx of unverified conspiracy theories, extreme content, and fake news, some of which were even propagated by Musk himself.
To combat the flood of misinformation, Musk implemented a "Community Notes" feature, allowing X users to prominently clarify facts or add background information below problematic posts.
Strong criticism of the Grok retraining plan
Musk's proposal has drawn condemnation from numerous critics, including AI startup founders and NYU neuroscience professor Gary Marcus, who likened the billionaire's plan to a dystopian world.
"This is like a reimagining of '1984'," Marcus commented on X. "You can't get Grok to align with your personal beliefs, so you decide to rewrite history to fit your viewpoint."
Bernardino Sassoli de’ Bianchi, a professor of logic and philosophy of science at the University of Milan, stated on LinkedIn that he is "at a loss for words regarding the dangers of Musk's plan."
"When powerful billionaires treat history as malleable simply because the outcomes do not align with their beliefs, we are no longer facing innovation—but narrative control," he added. "Rewriting training data to match ideology is wrong on any conceivable level."
As part of the effort to reshape Grok, Musk has called on X users to share "controversial facts" to train the bot, specifically noting that these facts should be "politically incorrect but still factually true."
Responses included various conspiracy theories and debunked extreme claims, including distortions of the Holocaust, discredited vaccine misinformation, racist pseudoscience regarding intelligence, and climate change denial.
Related: Reports indicate that the Bank of Korea's governor does not oppose issuing a won-pegged stablecoin.
Original article: “Musk Wants Grok AI to ‘Rewrite the Entire Corpus of Human Knowledge’”
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
