Elon Musk says that his artificial intelligence company XAI will resume his AI model, Grok, in a new knowledge base of “garbage” and “non -corrected data”, when using it first to rewrite the story.
In an X post on Saturday, Musk said that the next Grok 3.5 model will have “advanced reasoning” and wanted it to be used “to rewrite the entire corpus of human knowledge, adding missing information and eliminate errors.”
He said that the model would like to entertain again in the new knowledge set, claiming that there was “too much garbage in any base -trained base model.”
Musk’s last fight against “Wake”
Musk has long affirmed that the models of the rivals, such as the Openai chatgpt, a firm that co -founded, are biased and omit information that is not correct policy.
For years, Musk has sought to shape the products to be free of what Damag’s political correction considers and has aimed to make Grock what he calls “anti-woke.”
Hello, it also relaxed the Twitter content and the moderation of erroneous information when it delivered in 2022, which saw the flooded platform of non -controlled conspiracy theories, extremist content and false news, some of which extended by Musk itself.
Musk aimed to combat the tide of erroneous information through the implementation of a characteristic of “community notes”, which allows users of discrediting or adding context to the publications shown prominently under offensive publications.
Criticism was leveled in Grok’s releasing
Musk’s position attracted the condemnation of his critics, including Gary Marcus, a founder of the startup and professor at the University of New York Emeritus of Neuronal Sciences that compared the multimillionaire plan with a dystopia.
“Directly from 1984,” Marcus wrote in X. “You could make Grok align with your own personal beliefs so that you will rewrite the story to do so according to your views.”
Bernardino Sassoli of ‘Bianchi, professor of philosophy of logic and scientific at the University of Milan, wrote on LinkedIn that it was “with a loss of words to comment on how dangerous is Musk’s plan.
Related: Do you need Chatgpt help? I could be doing more dumb
“When the powerful billionaires deal with history as mixable simply because the results do not align with their beliefs, we are no longer dealing with innovation, we face narrative control,” he added. “Rewrite training data so that they coincide with ideology is incorrect at all conceivable levels.”
Musk’s call for “facts” brings conspiracy theories, false hoods
As part of his effort to review Grok, Musk asked X users to share “divisive facts” to train the bot, specific life should be “politically incorrect, but a meaningless reality.”
The answers saw a variety of conspiracy theories and discredited extremist claims, including the distortion of the holocaust, the evil iso of the discredited vaccine, the racist pseudoscientific statements with respect to intelligence and denial of climate change.
Ai Eye: The AI is good for employment, says PWC – Ignore the doomers of AI