Elon Musk founded a company called “X.AI” in Nevada in March 2023 to use the current hype about artificial intelligence (AI) positively. In particular, Musk wants to counter the successful AI chatbots “ChatGPT” and “Bard” by developing a “maximum truth-seeking artificial intelligence” called “TruthGPT“. This is supposed to differ from the other chatbots by understanding the nature of the universe and not delivering “politically correct” results like the other chatbots and thus lying. Musk, who was himself a co-founder and main investor of ChatGPT, is of course referring to the many Filters used by ChatGPT, for example not to output discriminatory or insulting content.However, ChatGPT is also supposed to follow a political ideology and thus falsify the answers.

Musk criticized in an interview that ChatGPT is programmed to provide false information and that artificial intelligence as a whole has the potential to destroy civilization. He wants to create a “third option” in artificial intelligence alongside Microsoft Azure and Google Bard. A few weeks ago, Musk joined an appeal by experts in an open letter calling for a general suspension of the development of AI technologies. The development break should be used to improve the regulations for AI so that their use is safe.

Large corporations such as Google, Microsoft and Meta have been working on AI programs for years, for example for translations, search queries and targeted advertisements. Musk said on Fox News that he used to be close friends with Google co-founder Larry Page. “He really seemed to want a digital super-intelligence, basically a digital god, as soon as possible,” Musk said. With AI, however, you “cannot just blast off and hope for the best”.

He is also concerned that the ability of ChatGPT and Bard and other AI programs to generate images based on text descriptions could be used to produce and spread misinformation. How ChatGPT works is explained: “It formulates texts by estimating word by word the probable continuation of a sentence, which sometimes leads to completely incorrect information being generated without the user being able to tell the difference.

Having already subjected ChatGPT to a series of tests, I have to say that he is not entirely wrong. Because the answers that are sometimes given out (if you read them carefully and question them) are sometimes make you shake your head. In any case, it is certain that at the latest with the launch of ChatGPT for the general public, the gold rush mood has been awakened among many investors and all companies that even remotely work with AI systems.