The AI chatbot war is in full swing

Anthropic, the Amazon- and Google-backed AI startup founded by former OpenAI employees, launched a new chatbot yesterday that it says outperforms ChatGPT-4, Google’s Gemini, and other similar products.

The new chatbot, called Claude 3 Opus, is able to understand text and photo inputs, though it lacks the ability to generate images. It officially launched yesterday for $20/month, with a less powerful version offered for free.

• Anthropic says Claude 3 Opus is particularly useful when analyzing scientific data or generating computer code – though it also outperformed the latest versions of ChatGPT and Gemini across every other category in a series of tests used to measure a chatbot's capabilities.
• These categories include math problem solving, undergraduate level knowledge, graduate level reasoning, and the ability to process visual inputs.
• Additionally, Claude 3 Opus can summarize up to ~150,000 words (aka an average Harry Potter book), while ChatGPT-4 can summarize ~25,000 words.

Anthropic is also emphasizing safe and responsible AI development. In an effort to help users feel they can trust their results, Anthropic says Claude 3 Opus will make every high school English teacher in America proud and start citing sentences from reference materials to back up its responses.

🤖💬 Looking ahead… Claude 3 Opus joins a wide range of existing AI chatbots on the market – some of which have differentiated themselves by specializing in certain areas. For example:

• When it comes to software coding, GitHub’s Copilot and ChatGPT-4 are widely considered the best available options.
• Gemini seamlessly integrates with nearly all of Google’s products (including Gmail, Drive, Meet, etc.).
• ChatGPT-4 and Microsoft’s Copilot – which is separate from GitHub – are among the handful of all-inclusive chatbots that can generate and edit images.