Anthropic's Claude 3 Outperforms GPT-4 in Benchmarks, Raises $5 Billion to Compete with OpenAI
-
Anthropic's new AI chatbot Claude 3 beats OpenAI's GPT-4 in performance tests, according to Anthropic's own benchmarks. Claude 3 is the first multimodal GenAI from Anthropic that can process images as well as text.
-
Claude 3 has increased capabilities for analysis, forecasting, and performance compared to ChatGPT, GPT-4, and Google's Gemini. It can also follow instructions better and produce structured outputs.
-
Claude 3 has a context window of up to 1 million tokens, allowing it to better grasp narratives and provide more contextually relevant responses. However, it still struggles with biases, hallucinations, and low-resource languages.
-
Anthropic plans frequent updates to Claude 3's capabilities, eventually aiming to create an "AI self-teaching" algorithm that could power advanced virtual assistants.
-
Anthropic aims to raise up to $5 billion to remain competitive with OpenAI. Its "constitutional AI" technique seeks to make models' behavior more predictable and adjustable.