Despite these limitations, Claude 3 has expanded context, allowing it to better follow multi-step instructions and produce structured output. It also has a more nuanced understanding of requests and will soon cite the source of its answers. Anthropic plans to release enhancements to the Claude 3 model family over the next few months. The company aims to create a next-gen algorithm for AI self-teaching, which could be used to build virtual assistants capable of answering emails, performing research, and generating art, among other tasks.
Key takeaways:
- AI startup Anthropic has announced the latest version of its GenAI tech, Claude 3, which it claims rivals OpenAI’s GPT-4 in terms of performance. The new GenAI includes models like Claude 3 Haiku, Claude 3 Sonnet, and Claude 3 Opus, with Opus being the most powerful.
- Claude 3 is Anthropic’s first multimodal GenAI, meaning that it can analyze text as well as images. However, it has limitations in identifying people and dealing with low-quality images or tasks involving spatial reasoning and object counting.
- Anthropic admits that Claude 3 isn’t immune from the issues plaguing other GenAI models, such as bias and hallucinations. The company promises frequent updates to Claude 3 in the coming months.
- Anthropic is planning to add features to Claude 3 that enhance its capabilities, including allowing it to interact with other systems, interactive coding, and more advanced agentic capabilities. The company aims to raise as much as $5 billion over the next 12 months to remain competitive with OpenAI.