Friday, 12 May 2023

“Incredible Achievement: OpenAI Outdone by Anthropic’s Chatbot that Reads a Novel in Less Than a Minute”

Anthropic’s AI startup has greatly expanded the context windows of its own chatbot Claude. This overcomes the limitation many chatbots face in their memory, as the amount they can process is restricted by their “context window” of input text and output. ChatGPT is limited to around 3,000 words, while Claude can now process 75,000 words, enough to encompass the entirety of The Great Gatsby. In fact, Claude’s system was tested by editing a single sentence in the novel and asking the chatbot to spot the change, which it did in 22 seconds.

It’s important to note that AI language models measure information not by the number of characters or words, but in tokens. A semantic unit that doesn’t map precisely onto these familiar quantities, tokens may be shorter or longer and do not always correspond to the complexity of meaning. Claude’s context window can now process 100,000 tokens, up from 9,000 previously. By comparison, OpenAI’s GPT-4 processes around 8,000 tokens, while a limited-release full-fat model of GPT-4 can handle up to 32,000 tokens.

Claude’s new capacity is currently only available to Anthropic’s business partners, who access the chatbot via the company’s API. While the pricing is unknown, processing more text means spending more on compute. Nonetheless, the news shows that AI language models’ capacity to process information continues to increase, making these systems more useful. With Claude’s expanded context window, it can potentially read, summarize, and analyze long documents in just a matter of minutes. A bigger context window also means the system is able to hold longer conversations, thus expanding the chatbot’s capabilities.

Editor Notes:

It’s exciting to see that AI language models are constantly expanding their potential. Anthropic’s achievement with Claude is remarkable, providing significant value for their current and future business partners. Hopefully, such advancements will improve the efficiency of chatbots, and we’ll see more widespread use of these systems in numerous industries. Check out GPT News Room for more exciting updates in AI technology.

Source link



from GPT News Room https://ift.tt/o10xhgS

No comments:

Post a Comment

語言AI模型自稱為中國國籍,中研院成立風險研究小組對其進行審查【熱門話題】-20231012

Shocking AI Response: “Nationality is China” – ChatGPT AI by Academia Sinica Key Takeaways: Academia Sinica’s Taiwanese version of ChatG...