Neowin®

Microsoft is reportedly working on its own AI large language model called MAI-1. It is being created to rival other LLMs from Google, Anthropic, and even ChatGPT from Microsoft's partner OpenAI.

Anthropic has launched a dedicated app for Claude on the App Store for iPhone devices. The app allows users to sync their chats with the web version as well as upload images for image analysis.

The UK Competition and Markets Authority has announced it is taking comments from others to see if the partnerships that Microsoft and Amazon have made with AI companies could harm competition.

AWS is giving startups up to $500,000 in credits for using different AI models by Anthropic, Meta, and more to be used on its own AI platform, Amazon Bedrock in partnership with Y Combinator.

Amazon has increased its funding for Anthropic to $4 billion. As part of this invesment, Amazon has a minority stake in Anthropic and Anthropic is using Amazon's cloud infrastructure.

Anthropic has released Claude 3 Haiku via its API, on claude.ai, on Amazon Bedrock, and soon on Google Cloud Vertex AI. Haiku is the fastest Claude 3 model, outperforms Gemini 1.0 Pro, and costs less.

Amazon Bedrock, a service that lets developers add generative AI to their applications, has added support for Claude 3 Sonnet. In the coming weeks, the Opus and Haiku editions will be added.

Anthropic has revealed its latest AI large language models. The Claude 3 family of LLMs includes its most advanced model, Claude 3 Opus, which the company claims beats GPT-4 and Gemini in benchmarks.

The Federal Trade Commission has just announced a new inquiry centering on AI partnerships and investments, with Microsoft, OpenAI, Google, Amazon, and Anthropic all being asked to send information.

Robin AI, a company which makes a Microsoft Word AI add-on to speed up work on legal documents, has just raised $26 million in funding to help it capitalize on its recent successes.

The Frontier Model Forum has shared its first working update and introduced an AI Safety Fund of more than $10 million. It has also appointed a new Executive Director to manage the forum.

The UK government has asked artificial intelligence companies to share more details of how their large language models work. It's in preparation for a global AI summit in the UK in November.

Amazon has made a major investment in Anthropic, taking a minority stake for $1.25 billion with an option for a total of $4 billion. Anthropic will use AWS as its primary cloud provider.

Anthropic has announced the availability of Claude Instant 1.2 through an API. Interestingly, benchmarks show it to be the safest of all the models so far with fewer hallucinations.

The four companies are teaming up to establish an advisory board that will help to create safe practices for "frontier AI models". More companies could join this group in the coming months.

Anthropic has released Claude 2, the latest version of its AI model. It's available for a fee via API or through a web chat if you are in the US or UK. The web chat lets you upload documents too.