Photo credit: www.investing.com
Safe Superintelligence Secures $1 Billion Investment
SAN FRANCISCO/NEW YORK – Safe Superintelligence (SSI), co-founded by Ilya Sutskever, the former chief scientist of OpenAI, has successfully raised $1 billion in funding aimed at creating safe artificial intelligence systems that aim to exceed human capabilities, as reported by company executives.
With a small team of 10 employees, SSI intends to allocate these funds towards enhancing its computing capabilities and recruiting elite professionals in the field. The company will maintain operations in both Palo Alto, California, and Tel Aviv, Israel, focusing on a tightly-knit team of adept researchers and engineers.
While SSI has opted not to disclose its valuation, insiders familiar with the situation have indicated that it stands at approximately $5 billion. This funding round highlights the ongoing interest from investors willing to place significant bets on exceptional talent and foundational AI research—despite a broader decline in enthusiasm for investments in unprofitable startups, which has seen many founders transition to larger tech firms.
Notable investors include major venture capital firms such as Andreessen Horowitz, Sequoia Capital, DST Global, and SV Angel, with additional participation from NFDG, an investment group led by Nat Friedman and SSI’s CEO Daniel Gross.
In an interview, Gross underscored the importance of aligning with investors who comprehend and support their mission of advancing safe superintelligence, stressing a commitment to several years of research and development before launching any products.
The conversation around AI safety has intensified, driven by concerns regarding potentially rogue AI systems that could compromise human interests or even threaten existence. In California, a legislative proposal aimed at establishing safety regulations has faced a divide in the tech industry, garnering opposition from companies like OpenAI and Google, while receiving backing from organizations such as Anthropic and Elon Musk’s xAI.
Sutskever, at 37, is regarded as a pivotal figure in artificial intelligence. He co-founded SSI in June alongside Daniel Gross, who previously spearheaded Apple’s AI initiatives, and Daniel Levy, a former researcher at OpenAI.
In the organization, Sutskever holds the position of chief scientist, with Levy serving as principal scientist and Gross overseeing computing resources and fundraising efforts.
A New Direction for Sutskever
Sutskever noted that his decision to start a new venture was prompted by the recognition of a different “mountain” to climb compared to his previous work.
Last year, he was part of the board overseeing OpenAI’s non-profit parent organization, which controversially voted to dismiss CEO Sam Altman due to communication failures. Following backlash, including a letter signed by nearly all OpenAI employees demanding Altman’s reinstatement, Sutskever quickly reversed his stance. His earlier actions, however, resulted in a diminished role within OpenAI, leading to his removal from the board and departure from the company in May.
After Sutskever’s exit, OpenAI disbanded his “Superalignment” team, whose goal was to ensure artificial intelligence remained consistent with human values, especially as AI begins to outperform human cognition.
Contrary to OpenAI’s unconventional corporate governance, aimed at safeguarding AI accountability, SSI has opted for a more traditional for-profit framework.
Team Composition and Culture
Currently, SSI prioritizes the hiring of individuals who align with its cultural values. Gross articulated their rigorous vetting process, emphasizing the importance of “good character” in potential hires, and expressing a preference for extraordinary talent over solely credential-driven evaluations. He remarked, “What excites us is discovering individuals who are genuinely passionate about the work, rather than those drawn by trends or hype.”
To meet its computational needs, SSI plans to forge partnerships with cloud service providers and semiconductor companies, although they have yet to finalize any specific collaborations. Many startups traditionally align with tech giants like Microsoft and Nvidia for their infrastructure requirements.
Sutskever has been a proponent of the scaling hypothesis, which suggests that increased computational power directly correlates with improved AI model performance. This principle has spurred substantial investment in the necessary infrastructure, ultimately fueling advancements in generative AI applications like ChatGPT.
Looking ahead, Sutskever indicated he plans to tackle scaling from a unique perspective compared to his previous work, although he refrained from elaborating on specific strategies.
“The discussion around scaling often focuses on the notion itself, yet frequently overlooks the critical question: what are we actually scaling?” he stated. He further emphasized that while some may work longer hours adhering to conventional paths, SSI’s approach seeks to foster innovation by exploring different methodologies.
Source
www.investing.com