[ad_1]
Two days after announcing the formation of xAI through a tweet, Twitter executive chair and CTO Elon Musk held a live audio session to talk about the big artificial intelligence project and the reason it was created, despite being vocally against it in the past. During the Twitter Spaces chat, Musk said that he is planning to “build a good AGI” or artificial general intelligence. He also highlighted that, unlike most AI platforms, xAI will be used to understand “what the hell is really going on”.
Expanding his thoughts, Musk explained that xAI will be aiming to solve complex scientific and mathematical problems in the world and through them, will try to understand the universe better. At one point he even raised the question, “Why is there not massive evidence of aliens?”, giving a hint at the kind of things xAI can be used to explore. How it may explore them, however, was not answered.
Another interesting take on the reason behind the creation of the company came from cofounder Kyle Kosic, who formerly worked for OpenAI. He said, “The reason I’m particularly excited about xAI is that I think the biggest danger of AI is monopolization by a couple of entities,” hinting that the company will try to increase competition in the AI sector.
Musk himself admitted that xAI is “definitely in competition” with OpenAI, the company behind GPT technology. Interestingly, Musk himself was an early investor in the company.
xAI to use data from public tweets for training
During the session, Musk also addressed the issue of the high volume of data scraping and alleged that “every AI organization on Earth” has illegally used the data from Twitter to train its models.
He explained why adding the Twitter rate limit on viewing tweets became necessary for the company. He said, “We had multiple entities scraping every tweet ever made, and trying to do so in like, basically a span of days. So, this was bringing the system to its knees. So we had to take action”. He also proceeded to apologize for the inconvenience caused due to the rate limit.
Musk also revealed that xAI will be trained on public tweets. He said, “I guess we will use the public tweets, obviously not anything private, for training as well. Just like basically everyone else has”. He added that Twitter’s data is appealing for text training as well as image and video training.
Not a “woke” chatbot
Musk highlighted during the session that the eventual generative AI that will be released by xAI will not be “politically correct”, hinting at popular chatbots such as OpenAI’s ChatGPT, Microsoft’s Bing AI, and Google’s Bard.
“I think our AI can give answers that people may find controversial even though they are actually true,” he said.
[ad_2]
Source link