Anthropic, an AI company, has launched a new chatbot called Claude 2 that aims to rival OpenAI’s ChatGPT. The updated version of Anthropic’s language model is capable of analyzing entire books and maintaining conversations for extended periods. Initially available to users in the US and the UK, the chatbot will expand to more countries soon. Users can access Claude 2 through Quora’s chatbot service, Poe.
Chatbots are AI applications that mimic human conversation to provide text-based responses on various topics. OpenAI’s ChatGPT gained significant popularity, accumulating 100 million users within two months. This success led to the emergence of other competitors like Microsoft’s Bing and Google’s Bard.
However, the rapid rise of chatbot technology has raised concerns among experts and lawmakers regarding privacy and societal risks. Italy, for instance, banned the use of ChatGPT in March due to data safety concerns. Many influential figures in the tech industry have called for a halt on advanced AI systems, asserting that these systems could pose an existential threat to humanity.
To address some of these concerns, Anthropic claims that Claude is built with trust and safety practices derived from major organizations and corporations. The chatbot was trained using documents such as the UN Declaration of Human Rights and Apple’s terms of service, among others. Anthropic has been working to enhance the safety of Claude 2, making it less likely to generate offensive or dangerous output.
Dario Amodei, the CEO of Anthropic, participated in a meeting with British Prime Minister Rishi Sunak in May, alongside other AI leaders. During the meeting, they discussed the potential risks of AI, ranging from disinformation and national security to existential threats. Amodei was among the tech experts who signed a statement emphasizing the importance of treating AI with the same urgency as nuclear war, as it could potentially lead to the extinction of humanity.
In May, Anthropic extended Claude’s capabilities by increasing the length of input and output. This allows the chatbot to analyze hundreds of pages of technical documents or even entire novels. Claude can also produce longer texts, including memos, letters, and short stories. Additionally, the chatbot can access up-to-date information from the web, a feature limited to paying users of ChatGPT.
Similar to ChatGPT, Claude boasts some additional advantages. Anthropic has improved the chatbot’s coding skills, with Claude achieving a score of 71.2% (up from 56%) on a popular Python coding test. It also scored 88% (up from 85.2%) on a comprehensive set of math problems from US primary schools. These enhancements further solidify Claude’s capabilities.
Anthropic aims to provide users with a conversational AI tool that offers valuable and reliable assistance. By combining extensive language comprehension, coding skills, and integration with various platforms, Claude 2 aims to deliver a high-quality chatbot experience to users worldwide.