Elon Musk continues building 'TruthGPT' despite cautioning on AI risks

Elon Musk has announced his intention to create a "TruthGPT" as an alternative to OpenAI's ChatGPT, which he believes has a liberal bias. 

The billionaire tech giant claims that his new AI will be a "maximum truth-seeking" system that aims to comprehend the universe's nature. He has criticised OpenAI for training its chatbot to be politically correct, despite being an early investor in the company. 

Musk's announcement comes just weeks after he called for a halt in the development of powerful AI systems, citing the profound risks they pose to society. ChatGPT is a large language model that interprets text and produces human-like responses after being trained on a vast amount of data. 

Its latest update, GPT-4, has significantly improved its ability to understand and complete various tasks, including writing poetry, solving computer bugs, passing exams, and assisting users with online searches and holiday bookings, with companies like Microsoft and Expedia integrating it into their products.

Models like ChatGPT and Google's Bard have occasionally been known to confidently spout outright falsehoods, raising concerns about the accuracy of the information they provide. This has led to calls for regulation as AI technology develops at a rapid pace. Elon Musk has expressed support for government action, believing the technology to be more dangerous than the rockets and cars his companies are known for. 

Tesla's self-driving cars have faced increased scrutiny, with over 360,000 vehicles recalled earlier this year due to concerns about their software causing accidents. Recently, the Chinese government published draft rules suggesting that generative AI services should adhere to the same tight restrictions as the rest of the internet in China. 

Italy banned ChatGPT in March while investigating its collection of user data, and EU law enforcement agency Europol warned that the model could be used to spread disinformation and be exploited by criminals. 

The White House has also unveiled its blueprint for an AI bill of rights to protect jobs and privacy and is seeking public feedback on how AI should be regulated in the US.

When will 'TruthGPT' launch?

Recent reports indicate that Musk is in the process of assembling a team for his AI start-up. However, he has already established a new company named X.AI Corp, according to a business filing in Nevada. 

The company was formed on March 9th and lists Musk as the director, with his adviser Jared Birchall serving as the secretary. This development follows the news that Musk had merged his Twitter account into a company known as X Corp.