New AI Chatbots by Anthropic Surpass GPT-4
Anthropic, an AI startup supported by Google and substantial venture capital funding, has unveiled the latest iteration of its GenAI technology, named Claude. The company asserts that Claude 3 surpasses the performance of OpenAI’s GPT-4 in the realm of AI chatbots.
Comprising models such as Claude 3 Haiku, Claude 3 Sonnet, and Claude 3 Opus (with Opus being the most potent), the GenAI family demonstrates “increased capabilities” in analysis and forecasting, according to Anthropic. The performance enhancements are claimed to outshine models like ChatGPT and GPT-4 (excluding GPT-4 Turbo) and Google’s Gemini 1.0 Ultra (excluding Gemini 1.5 Pro) on specific benchmarks.
Claude 3 marks Anthropic’s inaugural foray into multimodal GenAI, allowing it to analyze both text and images, akin to certain variants of GPT-4 and Gemini. This capability enables Claude 3 to process diverse visual data, encompassing photos, charts, graphs, and technical diagrams extracted from PDFs, slideshows, and other document formats.
An advanced feature sets Claude 3 apart from some GenAI competitors, as it can analyze multiple images in a single request, up to a maximum of 20. This functionality empowers the model to compare and contrast images, offering a notable advantage, as highlighted by Anthropic.
However, certain limitations accompany Claude 3’s image processing capabilities. Anthropic has deliberately disabled the model’s ability to identify individuals, evidently mindful of ethical and legal concerns. The company also acknowledges that Claude 3 exhibits challenges with “low-quality” images (those under 200 pixels) and faces difficulties in tasks requiring spatial reasoning, such as reading an analog clock face, and object counting—Claude 3 is unable to provide exact counts of objects in images.