Introducing Claude: Anthropic's Large Language Model

Anthropic introduces Claude and Claude Instant, reliable and safe "Constitutional AI," assistant, integration with various partner products in different sectors.

Introducing Claude: Anthropic's Large Language Model

Anthropic Debuts Claude and Claude Instant

On Tuesday, Anthropic unveiled its large language model (LLM), Claude, which can generate text, write code, and function as an AI assistant akin to ChatGPT. Claude stems from core concerns about the future of AI safety and has been trained using a technique Anthropic dubs "Constitutional AI."

A New Era of AI Assistance

Based on Anthropic's research into helpful, honest, and harmless AI systems, Claude is accessible via chat interface and API in the developer console. It is designed to handle a wide range of conversational and text processing tasks with exceptional reliability and predictability. Its potential applications include summarization, search, creative and collaborative writing, Q&A, coding, and more. Early users have praised Claude for being less likely to produce harmful outputs, easier to engage with, and highly steerable, resulting in desired outputs with minimal effort.

Limited Early Access and Commercial Partnerships

Claude and its counterpart, Claude Instant, are now available for a limited early access group and to commercial partners of Anthropic. Users can access Claude through a chat interface in Anthropic's developer console or via an application programming interface (API). The API allows developers to remotely connect to Anthropic's servers and incorporate Claude's analysis and text completion capabilities into their applications.

Advantages of Claude Over Other AI Chatbots

Anthropic asserts that Claude is "much less likely to produce harmful outputs, easier to converse with, and more steerable" than other AI chatbots, all the while maintaining "a high degree of reliability and predictability." Claude's use cases include search, summarization, collaborative writing, and coding. Like ChatGPT's API, Claude can alter its personality, tone, or behavior based on user preferences.

Pricing and Comparisons to OpenAI's ChatGPT API

Anthropic is pricing Claude based on usage per million characters input and output. While OpenAI's gpt-3.5-turbo AI model costs $0.002 per 1,000 tokens (fragments of a word), Claude Instant is priced at $0.42 per million characters for prompt input and $1.45 per million characters for output. The larger model, "Claude-v1," is available at $2.90 per million characters input and $8.60 per million characters output. Although there is no standard conversion between tokens and characters, OpenAI's ChatGPT API is roughly estimated to cost around $0.40 to $0.50 per million characters, making Claude generally more expensive.

Integration with Partner Products

According to Anthropic, Claude has already been integrated into several products available through partners, such as DuckAssist instant summaries from DuckDuckGo, a portion of Notion AI, and an AI chat app called Poe developed by Quora.

Quora: Fostering Natural Conversations with AI

Quora has implemented Claude in their AI Chat app, Poe, which users have found to be highly conversational, interactive, and creative in storytelling. According to Autumn Besselman, Head of People and Comms from Quora, users appreciate Claude's detailed and easily understood answers, as well as the natural flow of conversation.

Juni Learning: Revolutionizing Online Education

Juni Learning, a leading provider of online education solutions, has adopted Claude in their Discord Juni Tutor Bot. CEO Vivian Shen explains that Claude was chosen for its high-quality responses and ability to deliver a true tutor-like conversational experience, providing richer answers across a variety of subjects.

Notion: Enhancing Productivity and Writing Skills

Notion has integrated Claude into their workspace to improve users' writing abilities and increase productivity. Co-Founder & COO of Notion, Akshay Kothari, emphasizes that Claude's creative writing and summarization talents align with their goal of helping individuals and enterprises leverage AI.

DuckDuckGo: Reinventing Search with AI

DuckDuckGo has partnered with Anthropic to incorporate Claude in DuckAssist, an Instant Answer feature that uses natural language technology to generate answers from sources like Wikipedia. Steve Fischer, Chief Business Officer at DuckDuckGo, values Anthropic's commitment to improving DuckAssist while meeting strict privacy requirements.

Robin AI, a legal infrastructure business, is leveraging Claude to reimagine the future of contracts. CEO Richard Robinson reports higher user engagement, stronger feedback, and more closed deals since deploying Claude to evaluate contract language and offer customer-friendly alternatives.

AssemblyAI: Innovating Digital Media

AssemblyAI, an AI company specializing in audio data transcription and understanding, is partnering with Anthropic to enhance its platform. Founder & CEO Dylan Fox believes Anthropic's commitment to AI integrity and research will accelerate their development of robust, LLM-backed Generative AI and Conversation Intelligence capabilities.

Anthropic's Origins and Google Investment

Anthropic was founded in 2021 by former OpenAI VP of research Dario Amodei and his sister Daniela, following disagreements over OpenAI's increasingly commercial focus. Dario Amodei brought along other OpenAI staff, including Tom Brown, who led engineering work on GPT-3. A revision of GPT-3 later served as the foundation for ChatGPT. In late 2022, Google invested $300 million in Anthropic, acquiring a 10% stake in the company.

Google's Strategic Interest and Anthropic's Future Plans

Although Google has announced plans for Bard and the PaLM API, it seems unlikely that the tech giant will rely on Anthropic for AI solutions in its products. However, funding a rival to OpenAI may be in Google's strategic best interest in the long run. Anthropic plans to continue developing and improving Claude, with the company stating, "We plan to introduce even more updates in the coming weeks. As we develop these systems, we'll continually work to make them more helpful, honest, and harmless as we learn more from our safety research and our deployments."

The introduction of Claude could indeed mark the beginning of a new era in AI assistance, and I look forward to witnessing the impact it will have across various sectors.

Read next