Elon Musk, the billionaire, recently announced the development of a new artificial intelligence (AI) platform named ‘TruthGPT’. The platform is introduced as an opposition to Microsoft and Google’s products. He acccuses the Microsoft-backed OpenAI, the creators of ChatGPT, of “training the AI to lie” and stated that it has evolved into a “for-profit” closed source project “closely allied with Microsoft”. Additionally, Musk has rebuked the seriousness given to AI safety by Google co-founder, Larry Page.
TruthGPT is designed to be a “maximum truth-seeking AI” as opposed to OpenAI’s chatbot being trained to lie. He believes that this would be the safest path to take, as it is “unlikely to annihilate humans” as opposed to powerful AI systems.
In order to develop the TruthGPT, Musk has recruited AI experts from Alphabet Inc’s Google. This move is made despite a previous plea from AI researchers and industry leaders including Musk to stay clear of constructing systems more powerful than OpenAI’s GPT-4, citing potential dangers.
To further his cause of AI safety, he has created a company called X.AI Corp in Nevada last month, with himself as the only director and Jared Birchall, the managing director of Musk’s family office, as the secretary. Moreover, he also met with former US President Barack Obama, urging him to “encourage AI regulation” during his tenure.
Musk has been engaging in the field of AI for some time now. He co-founded OpenAI in 2015 and left the company in 2018, citing conflicting interests with Tesla and SpaceX.
Twitter was part of his investments as well, although he specified that it was valued at “less than half” of the acquisition price when Microsoft Corp decided to invest in OpenAI.
X.AI is a newer addition to the tech industry, although its purpose and plans remain shrouded in mystery. If Elon Musk’s intentions to create a maximum truth-seeking AI are successful, it will provide the tech industry and countless others with the foundational components to craft a viable, and reliable, AI program.