A California-based AI research startup, Anthropic, aims to raise up to $5 billion over the next two years to challenge OpenAI, the AI research lab founded by Elon Musk, and enter into a dozen industries. TechCrunch managed to acquire copies of a pitch deck document that revealed such long-term ambitions of the company, co-founded in 2020 by former OpenAI researchers.
In the pitch deck, Anthropic gave details of its plan to construct a quality model, tentatively called “Claude-Next,” which would be 10 times more productive than the current powerful AI models. The project requires $1 billion to be spent across 18 months. At the time the document was published, it was reported that Anthropic was seeking $300 million at a $4.1 billion valuation, totaling up to $1.3 billion by the time the documents were made. This data was also revealed in the pitch deck.
The Claude-Next model is described as a “next-gen algorithm for AI self-teaching.” It is based on a technique called “constitutional AI,” which works towards making AI systems more aligned with human intentions. These AI systems, when trained with constitutional AI, can respond to questions and carry out tasks with a straightforward set of principles.
Anthropic estimated the complexity of the model at 10^25 FLOPs (Floating Point Operations) — an astronomical number compared to current models. To process this powerful model, the company likely relies on computers operated by tens of thousands of GPUs.
More practically, the model can be used to create virtual assistants that can answer emails, compile research, write books, and generate artwork — features which exist in AI models today, like the OpenAI’s ChatGPT. Anthropic believes that companies that produce the best 2025/26 model will be difficult or impossible for others to catch up to afterward.
The “founder” of Anthropic is Dario Amodei, former VP of research at OpenAI. His departure from OpenAI was due to differences in expectations and concepts, as OpenAI had a more business-oriented focus. Alameda Research Ventures and Google are also included in the list of investors, with the former reportedly spearheading the $580 million Series B round, estimated at a 10% stake.
In March 2021, Anthropic released its chatbot, Claude, commercially and has taken customers from a variety of industries, such as technology, finance, logistics, healthcare, and more, where the model is supervised by a human. It is claimed to be more controllable and challenging-to-produce harmful outputs compared to other AI models of today.
Anthropic’s series C fundraising has a goal to reach the limit of $5 billion, and with a larger-than-life model in the works, it is a project that has the potential to dominate the industry and make significant strides in the progress of AI research and its real-life applications.
