Highlights:

  • SSI aims to create highly advanced AI models with superintelligence, incorporating safeguards to prevent harmful outcomes.
  • Sakana, a generative AI startup founded in Tokyo a year before SSI, develops cost-efficient, small language models using nature-inspired machine learning techniques.

An AI startup founded by OpenAI’s former chief scientist, SSI, has secured USD one billion in funding to advance its research.

The company recently revealed the investment, which follows funding milestones from two other generative AI startups: enterprise chatbot provider You.com Inc. and language model developer Sakana AI Co. Ltd. Nvidia Corp. took part in the funding rounds for both companies.

Secure AI

SSI, formally known as Safe Superintelligence Inc., was founded in June by Ilya Sutskever, former Chief Scientist at OpenAI. He co-launched the startup with Daniel Levy, a former OpenAI researcher, and Daniel Gross, a former Y Combinator partner. Gross, who serves as SSI’s CEO, previously led Apple Inc.’s machine learning group earlier in his career.

SSI aims to create highly advanced AI models with superintelligence, incorporating safeguards to prevent harmful outcomes. At the time of the startup’s launch in June, Sutskever stated in an interview that SSI’s sole focus was on developing these models, indicating that the company has no immediate plans to introduce commercial products.

The recent announcement of SSI’s over USD one billion investment featured contributions from Sequoia, DST Global, Andreessen Horowitz, and SV Angel. Also participating was NFDG, an investment entity led by SSI CEO Daniel Gross and former GitHub CEO Nat Friedman. According to sources, the funding round, which reportedly values SSI at USD five billion, will be allocated to hiring and infrastructure development.

Nature-driven AI Models

Sakana, a generative AI startup that also secured funding recently, was established in Tokyo about a year prior to SSI. It focuses on developing small language models optimized for cost efficiency. The company claims its algorithms are based on machine learning techniques inspired by natural processes.

In March, Sakana introduced one of its nature-inspired techniques called evolutionary model merge, which involves combining open-source models to create improved versions. Using this method, the company developed a small language model capable of understanding Japanese text and solving math problems. With 7 billion parameters, the model reportedly outperformed language models ten times its size in internal tests.

Sakana recently announced a Series A funding round valued at over USD 100 million. The investment was led by New Enterprise Associates, with contributions from Khosla Ventures, Lux Capital, and Nvidia. The funds will be used to acquire infrastructure, support research initiatives, and foster the AI ecosystem in Japan.

Efficiency-enhancing Chatbots

Nvidia’s name surfaced in another funding announcement recently. You.com disclosed that it secured USD 50 million from investors, including the chipmaker Salesforce Ventures, DuckDuckGo, SBVA, and Day One Ventures. Georgian led the Series B round, bringing the startup’s total funding to USD 99 million.

You.com offers a cloud-based service featuring over half a dozen AI chatbots. One of these, promoted as a research assistant, can analyze multiple sources to respond to user queries. Another chatbot is designed to generate Python code from natural language prompts.

Users who find the built-in chatbots insufficient for their needs can create their custom versions. Alongside its funding announcement, You.com launched new collaboration features that enable workers to share their custom chatbots with team members. These features are being introduced for the platform’s paid Team tier.

You.com reports a fivefold increase in its annual recurring revenue since the beginning of the year, driven by strong demand. The platform is now utilized by publicly traded companies, venture-backed startups, and other organizations.