July 2024 Summaries

Anthropic partners with Menlo Ventures to launch Anthology Fund, a $100 million initiative to accelerate the development of groundbreaking AI applications in various industries such as healthcare, education, energy, infrastructure, and scientific research. The Fund will support startups innovating in key areas including AI infrastructure, novel applications of AI, consumer AI solutions, trust and safety tooling, and AI apps that maximize societal benefits. Startups backed by the Anthology Fund will gain access to Anthropic products and research, as well as venture support from Menlo Ventures. The partnership aims to accelerate the development of groundbreaking AI applications and drive responsible AI innovation.

The new Claude Android app allows users to access the power of Claude, including its most powerful model, Claude 3.5 Sonnet, on their Android devices. The app is free and compatible with all plans, including Pro and Team. It offers multi-platform support, vision capabilities, multilingual processing, and advanced reasoning, making it suitable for both work and personal use. Users can access Claude from anywhere, whether they're drafting documents or composing speech, and can download the app on Google Play to get started.

Customers can fine-tune Claude 3 Haiku in Amazon Bedrock to customize its knowledge and capabilities for their business, making the model more effective for specialized tasks. Fine-tuning allows companies to acquire specialized business knowledge, leading to improved accuracy and consistency, while reducing costs and improving speeds. The technique is easy to use without extensive AI expertise required. Companies such as SK Telecom and Thomson Reuters have seen positive results with fine-tuned Claude 3 Haiku models, including improved customer experiences and faster AI results. Fine-tuning for Claude 3 Haiku in Amazon Bedrock is now available in preview, and companies can request access through their AWS account team or support ticket.

To streamline the process of crafting high-quality prompts for AI-powered applications, Anthropic has added features to its console that allow users to generate, test, and evaluate their prompts more efficiently. The new features include a built-in prompt generator powered by Claude 3.5 Sonnet, automatic test case generation, and comparison of model responses. These enhancements enable developers to refine their prompts quickly and make data-driven decisions to improve the performance of their applications. With these improvements, users can reduce manual effort and increase confidence in the quality of their prompts, ultimately leading to better outcomes for their AI-powered projects.

A new initiative has been launched to fund third-party model evaluations, aiming to develop high-quality safety-relevant evaluations for assessing AI capabilities and risks. The focus areas include AI Safety Level assessments, advanced capability and safety metrics, and infrastructure tools for developing evaluations. Evaluations will be prioritized based on their ability to measure real-world risks accurately, with a particular emphasis on cybersecurity, CBRN risks, model autonomy, social manipulation, misalignment risks, advanced science, harmfulness and refusals, multilingual capabilities, societal impacts, and infrastructure development. The initiative aims to provide valuable tools that benefit the entire AI ecosystem, with a focus on collaboration and iteration to refine evaluations for maximum impact.