During the presentation of his artificial intelligence startup called xAI, Elon Musk offered little indication of how it works. However, he made it clear that this tool will not be trained to be politically correct.
Musk mentioned that xAI aims to compete with OpenAI, Google and Microsoft. However, during the audio-only presentation, the controversial billionaire did not provide specific information about the tools that xAI will offer, the start date of the services, who will have access to it, or whether it will be free to access.
He didn’t do any visual demonstration of how xAI will be able to work.
However, yes made a comparison with the most popular generative Artificial Intelligence (AI) chatbots, such as ChatGPT, Bard or Bing, They don’t respond to user questions if they think they’re racist, violent, or hateful, which Musk says shows they’re programmed to be politically correct.
“I think there’s a significant danger in training an AI to be politically correct or, in other words, training it not to say what it actually thinks to be true,” Musk noted in the company’s presentation he gave today at the Twitter Spaces chat, a social network that he has also owned since last October.
To clarify his opinion, the executive director of SpaceX and Tesla gave the example of the Stanley Kubrick movie “2001: A Space Odyssey”, and said that in that adventure things began to go wrong when they asked the supercomputer Hal 9000 to lie .
The xAI team would create a safer AI
The billionaire businessman took advantage of this coming-out to introduce the eleven men (all were varnos) behind xAI, several of whom have worked for leading companies in the sector such as OpenAI, DeepMind, and Microsoft.
Musk and his team stressed on several occasions that AI regulation will be key for them, but they did not detail how it will be carried out specifically, especially when this regulation is still in its infancy despite the numerous voices calling for some kind of international code to be implemented.
Other pillars of xAI will be to create a safer AI, “that is as curious as possible and seeks the truth”, since the ultimate goal is to understand the true nature of the universe.
In the presentation, Musk took questions from several Twitter Blue users, all men, including hacker Kim Dotcom; the writer of the biography of Steve Jobs and the book entitled “Elon Musk” that will go on sale next September, Walter Isaacson, or the expert in space computing Robert Scoble.
In one of the answers, Musk explained that xAI will use public tweets from Twitter to train and stressed that this is something that all big and small AI companies have done illegally.
ChatGPT became a “greedy” company
Musk described OpenAI, creators of ChatGPT, as a “greedy” company.
This company, which began collaborating with Microsoft this year, began as a non-profit artificial intelligence research company to “advance digital intelligence” and make it “benefit humanity as a whole, without the limitations of the need to generate financial returns” in December 2015.
Musk was one of the co-founders of OpenAI and rose to become co-chair, before leaving in 2018, a departure under unclear circumstances: OpenAI and Musk said at the time, South African left OpenAI to avoid conflicts of interest as Tesla it focused more on AI.
However, according to Semafor, Elon Musk had proposed to take over the leadership of OpenAI and withdrew after his proposal was rejected.
For its part, the Financial Times reported that Musk’s departure was also due to clashes with other board members and staff over OpenAI’s approach to AI security.
#Elon #Musk #summarizes #trained #politically #correct