Artificial Intelligence Research Society Anthropic Today announcement that he has raised $580 million in funding to support his work.
The Series B round was led by Sam Bankman-Fried, founder and managing director of cryptocurrency exchange FTX Trading Ltd. Caroline Ellison, Jim McClave, Nishad Singh, Jaan Tallinn and the Center for Emerging Risk Research also participated.
Anthropic is led by siblings Dario Amodei and Daniela Amodei, who were vice presidents of AI research lab OpenAI LLC before launching the company in 2020. Anthropic’s 40-person team is researching ways to build better neural networks. The company’s efforts have produced several academic papers since its launch two years ago.
One of Anthropic’s main areas of focus is AI security. The company is developing ways to develop large-scale AI models that have better safeguards and also provide the ability to verify the effectiveness of those safeguards. Additionally, the company is studying the societal impact of advances in machine learning.
Another area of research that Anthropic prioritizes is the area of scaling laws. As part of its work in this area, the company investigates how factors such as the configuration of an AI model, its performance and the data processed by the AI are related. Anthropic uses the patterns it finds to find more efficient ways to train neural networks.
“With this fundraiser, we will explore the predictable scalability properties of machine learning systems, while closely examining the unpredictable ways in which capabilities and security issues can emerge at scale,” said the co. -founder and CEO of Anthropic, Dario Amodei. . “We have made significant initial progress in understanding and driving the behavior of AI systems, and we are gradually assembling the elements necessary to create integrated and usable AI systems that benefit society. »
Anthropic plans to increase its workforce through the new round of funding. As part of this effort, the company is setting up dedicated teams and partnerships to research the political and societal impacts of new AI models.
“Now that we have built the organization, we are working to ensure that Anthropic has the culture and governance in place to continue to responsibly explore and develop safe AI systems as we scale” , said Anthropic co-founder and president Daniela Amodei.
Anthropic will also use the funding to deploy additional technology infrastructure that will help its team research computationally intensive neural networks.
Developing new neural networks can require a significant amount of hardware. OpenAI, for example, builds AI models using a supercomputer hosted on Microsoft Corp’s Azure cloud platform. which includes 10,000 graphics cards and 285,000 CPU cores. According to Microsoft, the system ranked among the five fastest supercomputers in the world when it was announced in 2020.