Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Igor Babuskin, co-founder XAI, during the Nvidia GPU Technology (GTC) conference in San Jose, California, USA, Tuesday, March 19, 2024.
David Paul Morris | Bloomberg | Gets the image
Igor Babuskin, a founder member Elon Musk XAI, said on Wednesday that it leaves the startup of artificial intelligence to launch its own venture firm.
“Today was my last day in XAI, the company I helped start with Elon Musk in 2023,” Babuskin wrote on Xi. “I still remember the day when I first met Elon, we talked about II for hours and what the future could happen. We both considered that the new AI campaign was needed. Building the II, which reached humanity, was my lifelong dream.”
Musk wrote in response: “Thank you for helping build @xai! We wouldn’t be here without you. “
Babuzhin said he begins Babuschkin Ventures to support AI security research and invest in “AI and Agentic Systems, which promote humanity and unlock the secrets of our universe.”
Former scientific engineer for Google’s Deepmind and former Openai technical staff, Babuskin told some of the major XAI prompt achievements during his stay, including the creation of engineering teams in the company.
“Through blood sweat and tears, the speed of our team built a supercracker of Memphis and supplied border models faster than any company in history,” he wrote.
The object in Memphis processes the data and prepares the models that Power XAI in Grok Chatbot.
Locals protested protest XAI operations in Memphis, especially the use of turbines that burn natural gas for their data centers. It is reported that the turbines are impaired by poor air quality in the city of Western Tennessee.
At a time when he was preparing to do business with Musk, Babuskin wrote that he believed that “very soon II could think for the level of people,” and was concerned that such technology is “used for the benefit”.
He said “Elon warned of the danger of powerful II for many years” and shared his vision “AI used to benefit humanity.”
The company has rocky services when it comes to AI safety.
In May, the Grok -chat -bot automatically created and distributed false reports of allegedly “white genocide” in South Africa. After that company apologized And said the crook Amazing behavior It was caused by “unauthorized modification” in Chatbot system clues, which help to report how it behaves and interacts with users.
In July, XAI was apologized for another problem with Grok. After upgrading Chatbot code automatically generated and distributed false and Anti -Semitic content Through x, including messages that concern Adolf Hitler.
Last month, the European Union asked for a meeting with XAI representatives to discuss XA problems and integrated Grok Chatbot.
Babuskin and XAI did not respond to comment requests.
Other chat -bots have also created false or other dangerous exits in response to requests. CHATPT OPENAI has recently been summoned for cottage Bad Health Parade The user that ended in intensive care. And Google had to do change For Gemini last year after it created offensive images in response to users’ clues, including images that reflected the colors of people as Nazis.
