Large Action Models

Discovered something way cool than ChatGPT – Large Action Models

I was little skeptical about this Rabbit R1, so I did some research on how the Rabbit R1 actually works.

So, Its a Device which runs a Large Action Model (LAM). So give you an idea of what LAM

You might have heard about LLM - Large Language Model which is used for Generative AI, the most famous ChatGPT runs on a LLM. LLM have made a significant impact across various sectors, enhancing chatbots, customer service, and providing valuable information based on user queries. The contributions to content creation, generating posts, emails, blog outlines, travel planning, recipe suggestions, research on specific topics, creating outlines, image processing, image and video generation, and much more. The possibilities are extensive, fostering continuous innovation.

The primary function revolves around predicting the next step based on input, utilizing different models and their training to yield results.

However, the natural progression involves taking actions, as while ChatGPT can offer plans, it cannot execute tasks like booking transport or placing meal orders.

Conventional personal assistant systems such as Alexa and Siri also fall short in this regard. This is where the concept of Large Action Models (LAMs) comes into play.

This introduces new realms of possibilities, envisioning a future where ChatGPT could potentially gain access to applications and perform tasks on behalf of users.

Here are some potential use cases:

• Booking a taxi
• Planning an entire trip, including transportation, accommodation, and flights, involving multiple websites or apps
• Real-time translation and audio/video conversations with accompanying actions
• Scheduling a meeting, sending invitations, and inviting relevant participants based on emails and calendar availability
• Analyzing social media posts to plan a content calendar and scheduling Tweets from a pool of generated blogs and content

The future, particularly in 2024, promises to see more developments in this field from hardware and phone manufacturers. In essence, the synergy of Large Language Models and Large Action Models holds immense potential, opening up futuristic possibilities and allowing users to save time for more crucial endeavors.

Large Action Models (LAMs), which involve AI systems capable of understanding and executing tasks based on natural language instructions, have significant potential applications in robotics, the Internet of Things (IoT), and other domains. In robotics, LAMs can enable robots to perform complex actions in response to spoken or written commands, enhancing human-robot interaction. In the IoT, LAMs can be used for smart home automation, industrial automation, and controlling various connected devices through natural language commands. The link between LAMs and these applications lies in their ability to make human-machine interaction more intuitive and user-friendly, leading to more efficient automation processes and improved user experiences.

Imagine what this tech could do when combined with Robotics, IoT or other applications.

Resources providing further insights:

• Rabbit Tech Keynote
• Salesforce AI Research Blog on Large Action Models