On the forefront of AI tech, especially those of the generative kind, are Google and OpenAI with its Microsoft partnership. Elon Musk, perhaps unsurprisingly considering his history with the latter, is not happy with these two being the major players in the field. So naturally, he has announced that he is working on one “TruthGPT”, which he dubs as an alternative.
During an interview with Fox News, Musk claims that his TruthGPT will be a “maximum truuth-seeking AI that tries to understand the nature of the universe”. The logic is apparently that an AI that understands the universe won’t go all Skynet on us humans “because we are an interesting part of the universe”.
In a sense, he is making good on a random, seemingly innocuous tweet he made back in February. But the whole endeavour does seem a bit hypocritical. After all, last month he signed an open letter asking AI companies to pause “giant experiments” in the field for six months. But at around the same time, there were reports of Musk hiring two former DeepMind researchers for an AI project.
More recently, he was reported to have started his own AI company called X.AI Corp. In the same time window, he was also found to have bought about 10000 GPUs for a generative AI project which, at the time, was thought to be for Twitter.
(Source: Fox News, Future of Life Institute, Business Insider, WSJ)
