The co-founder and ex-chief scientist of OpenAI, Ilya Sutskever, has launched a fresh AI venture called Safe Superintelligence (SSI). This company will utilize Google Cloud's Tensor Processing Unit (TPU) chips for advancing its AI studies as per an announcement made recently about their collaboration. The news was shared with the public on Wednesday. press release .
Google Cloud mentions that SSI is employing TPUs to "speed up its research and development activities aimed at developing a secure, highly intelligent AI system."
Cloud service providers are pursuing several high-value artificial intelligence start-ups that invest substantial sums, reaching hundreds of millions annually, in computational resources for training advanced AI models. According to reports, SSI has entered into an agreement with Google Cloud indicating that they plan to allocate a significant portion of their computing expenses through this platform. A reliable source disclosed to Maximathat Google Cloud serves as SSI’s main supplier of computational services.
Google Cloud has a track record of entering significant computational partnerships with ex-AI researchers from the company, numerous individuals who have since founded multibillion-dollar AI enterprises. (Sutskever previously held a position at Google.) In October, Google Cloud announced it would serve as the main computing provider for World Labs founded by former Google Cloud AI lead scientist Fei-Fei Li.
It remains uncertain whether SSI has formed alliances with additional cloud or computing service providers. When contacted, a representative from Google Cloud chose not to comment. Similarly, a response was not promptly received after reaching out to a Safe Superintelligence spokesperson for their input.
SSI emerged from stealth mode in June 2024, several months after Sutskever. departed from his role As the chief scientist at OpenAI, the company has received $1 billion in funding from investors including Andreessen Horowitz, Sequoia Capital, DST Global, SV Angel, and more.
Since the launch of SSI, we haven’t had much information about what the company has been up to. Its website SSI states that creating secure, highly intelligent AI systems is "our purpose, our identity, and our complete product strategy," as it represents their exclusive concentration. previously said He has recognized "a fresh peak to conquer" and is exploring innovative methods to enhance the efficiency of cutting-edge AI models.
Prior to launching OpenAI, Sutskever dedicated multiple years at Google Brain focusing on neural network research. Following his extensive involvement in overseeing OpenAI’s AI safety initiatives, Sutskever was instrumental in the removal of OpenAI CEO Sam Altman in November 2023. Subsequently, he became part of an employee initiative aimed at restoring Altman to his position as CEO.
Following the difficult experience, Sutskever apparently did not visit OpenAI’s offices for several months and eventually departed from the company to launch SSI.