Akihabara News (Tokyo) — Twitter owner and Tesla founder Elon Musk has had enough with lying chatbots.
Taking a shot at OpenAI and its ChatGPT during a recent Fox News interview, Musk declared that “they’re training the AI to lie.”
According to Musk, OpenAI programmers are not so open when it comes to providing answers and analysis which company officials deem to be not politically correct.
In order to combat what he sees as a troubling trend, Musk is advancing what he calls “TruthGPT,” an alternative platform which will be a “maximum truth-seeking AI.”
Musk has yet to provide any details about his project, but two months ago he founded a start-up company called X.AI.
Also, Musk has poached two researchers from the UK-based company DeepMind, which is specialized in AI programming; and he also purchased 10,000 graphics processing units (GPUs), which will convert data into graphics in an optimal manner.
Although he was a cofounder of OpenAI, Musk has long been wary about this technology’s downside risks. He has even gone so far as to suggest that, at some point, AI devices could outnumber humans, outsmarting and eventually replacing humanity.
He recently led an initiative calling for a six-month pause on new AI technology public rollouts in order provide government regulators a chance to legislate the necessary safeguards.
There is no indication, however, that such a pause will actually be implemented. Instead, ChatGPT is quickly growing in popularity with over 100 million active users.
Recent Artificial Intelligence Related Articles
Japan Grapples with Arrival of Chatbots