anthropic, an AI startup co-founded by former employees of OpenAI, has started offering its partners access to its AI text generation models.
The first commercial company to announce the integration of Anthropic models is Robin AI, a legal tech startup that has raised more than $13 million from investors including Plural, Episode 1, and the Google Black Founders Fund. Quora's experimental chatbot app for iOS and Android, Poe, uses anthropic models, but is currently not monetized.
Robin CEO Richard Robinson disclosed few details about the relationship with Anthropic, but said Robin worked to refine an Anthropic model on a dataset of legal text for drafting and negotiating contracts.
“We are very fortunate to be Anthropic's launch partner for the legal sector – the team's focus on AI security aligns with our 'lawyer-in-the-loop' software-as-a-service product – deliberately designed to manage the risk that even the most advanced models 'hallucinate'”Robinson said in a statement.
Anthropic has been relatively quiet about his plans to produce his work in the realm of generative text AI, preferring to focus on academic research. But late last year, the company launched a closed beta of an AI system, called Claude, similar to OpenAI's ChatGPT, which seemed to improve on the original in key ways. Claude was built using a technique developed by Anthropic called "constitutional AI," which aims to provide a "principles-based" approach to align AI systems with human intentions, letting ChatGPT-like AI answer questions using a simple set of principles (for example, avoid giving harmful advice) as a guide.
Claude's preliminary impressions were good. But, like ChatGPT, the system suffered from limitations, including giving dangerous answers to questions (such as how to make meth at home) and making inconsistent and misleading statements of fact.
It's unclear if the model Robin uses is Claude or some derivative: neither Robin nor Anthropic would say. And even after repeated prodding, Anthropic would not disclose how many partners it currently works with (or how they came to work with them) or how many models it plans to open up for commercial use.
But Anthropic is certainly feeling the pressure from investors to recoup the hundreds of millions of dollars invested in its AI technology.
Recently, Google has invested $300 million in Anthropic in exchange for a 10% stake in the company. Under the terms of the deal, first reported by the Financial Times, Anthropic agreed to make Google Cloud its "preferred cloud provider" and the companies will "co-develop AI computing systems."
Anthropic was not founded solely for profit. Dario Amodei, OpenAI's former vice president of research, launched the company in 2021 as a public benefit corporation, taking several OpenAI employees with him, including former OpenAI public policy lead Jack Clark. Amodei parted ways with OpenAI following a disagreement over the company's direction, specifically the startup's increasingly commercial focus.
But AI systems are expensive to develop and maintain. Rising costs prompted Anthropic to seek outside support, including a $580 million tranche from a group of investors including FTX founder Sam Bankman-Fried, Caroline Ellison, Jim McClave, Nishad Singh, Jaan Tallinn and the Center for Emerging Risk Research.
It's not yet clear whether the partnerships with startups and investments from big tech signal a shift in Anthropic's priorities. But what is clear is that the company believes its technology is differentiated to compete with rivals such as OpenAI, Cohere and AI21 Labs, all of which offer paid access to its text-generating AI via APIs.
