Anthropic, a startup supported by major tech players such as Microsoft, Google, Salesforce, and Zoom, has launched its latest AI chatbot called Claude 2. This move comes as part of the intense competition between Microsoft-backed OpenAI and Google to develop the most advanced chatbots powered by generative artificial intelligence. Anthropic, having raised $750 million in recent financing rounds, aims to keep pace with its rivals and capitalize on the growing demand for sophisticated chatbot technology.
With origins dating back to 2021 and co-founded by former OpenAI research executives Daniela Amodei and Dario Amodei, Anthropic is now opening its chatbot technology to consumers for the first time with the release of Claude 2. Over the past two months, Anthropic has been conducting tests with prominent businesses like Slack, Notion, and Quora, accumulating a waitlist of more than 350,000 individuals eagerly requesting access to Claude’s application programming interface and consumer offering.
Daniela Amodei expressed the company’s focus on enhancing Claude’s safety and robustness for businesses. However, she emphasized that Claude 2 represents the safest version developed thus far and is thrilled to offer it to a broader range of users, including both businesses and individual consumers. Initially, Claude 2 will only be accessible to users in the United States and the United Kingdom, but Anthropic plans to expand availability to more regions in the coming months.
The entry of Anthropic’s Claude 2 into the chatbot market comes on the heels of OpenAI’s introduction of ChatGPT and subsequent updates to its underlying technology. OpenAI’s GPT-4, a major advancement in chatbot capabilities, has garnered significant attention in the tech world. Likewise, Google launched its own chatbot, Bard, in February, incorporating AI advancements into various products.
Anthropic is determined to establish itself as a significant player in the field. Claude 2 demonstrates its ambition by offering impressive features, such as the ability to summarize up to approximately 75,000 words, equivalent to the length of a book. Users can input extensive datasets and request summaries in the form of memos, letters, or stories. In comparison, ChatGPT, developed by OpenAI, is limited to handling about 3,000 words.
Daniela Amodei affirmed that the demand for large language models is currently surpassing the industry’s ability to provide them. She mentioned that this is an exceptional period for businesses in the sector, with ample room for multiple players to succeed due to the wide landscape and diverse user requirements.
In May, Anthropic joined other prominent companies, including Google’s parent company Alphabet, Microsoft, and OpenAI, in a meeting at the White House to discuss responsible AI development with Vice President Kamala Harris. This recognition, coupled with recent successful funding rounds, including a $450 million raise in May and a $300 million financing round in March, has positioned Anthropic as a key player in the market.
Amodei provided tangible evidence of Claude 2’s improvements over its predecessor. The upgraded chatbot achieved a higher score on the Bar exam’s multiple-choice section, jumping from 73% to 76.5%. Moreover, in a Python coding test, Claude 2 scored an impressive 71%, compared to 56% in its prior iteration. The blog post revealed that the new chatbot also demonstrated an enhanced ability to provide “harmless responses” in safety evaluations.
Despite the progress made, Amodei acknowledged that challenges remain for Anthropic and the wider industry. One persistent problem is the tendency for AI chatbots to provide incorrect responses, often referred to as “hallucinations.” Amodei emphasized that no language model, including Claude 2, is completely immune to this issue.
Anthropic’s launch of Claude 2 marks a significant milestone in the race for advanced AI chatbot technology. With substantial financial backing and support from tech giants, Anthropic aims to cater to the soaring demand for powerful generative AI chatbots. While the company celebrates Claude 2’s enhanced performance and safety measures, it remains committed to addressing the remaining challenges and pushing the boundaries of language model development.