Anthropic, the maker of the popular Claude AI assistant, announced that Amazon will invest up to $4 billion into the company.
This will make Anthropic’s “safe and steerable” AI widely accessible to AWS customers.
AWS will also become Anthropic’s primary cloud provider for mission-critical workflows. (Importantly, that gives Anthropic access to Amazon’s chips.)
This will open up access to Claude 2 to many more organizations that are already building on AWS and want to access state-of-the-art models in safe, secure ways.
Anthropic cites some companies already building with Anthropic models via Amazon:
- LexisNexis Legal and Professional is using a custom version of Claude 2 to summarize and draft legal content.
- Famed investment firm Bridgewater Associates is using Claude 2 to generate charts and summarize data.
- And Lonely Planet is using Claude 2 to drop the cost of generating travel itineraries by 80% using all the decades of travel content they have.
Connecting the Dots
On Episode 66 of The Marketing AI Show, Marketing AI Institute founder and CEO Paul Roetzer talked me through what’s going on here.
- Start paying attention to Anthropic. “If you weren’t paying attention to Anthropic before, now would be a good time to start,” says Roetzer. Their technology is impressive, and they now have very, very deep pockets.
- If you’re an AWS customer…you can now start training Claude on your data safely and securely in a platform already approved by procurement.
- If you’re not an AWS customer…this type of functionality is being built into other major cloud providers. Check with yours to see which foundation models you can start building on.
- Building on top of these models is the future. “This is the future,” says Roetzer. “If your data lives in the major cloud providers (Google, Microsoft, Amazon), all three of them are trying to build the ability for you to develop language models in your company based on these foundation models.”
- And you’ll likely have and use many different models. “The assumption we’re under is you’re probably going to have a collection of these models in your company that are tuned specifically for different applications or different business functions,” says Roetzer. You may have a version for customer service, marketing, HR, sales, etc.
What to Do About It
- Find out who can help you solve for this. There are people internally (engineers, AI leaders, domain experts) and outside your organization (consultants) who can help you start solving for what models to build on and how to get the most out of foundation models in your organization. Find out who they are—and get the conversation started.