The Thousand Brains project: a new paradigm in AI that challenges deep learning with inspiration from the human brain

We have established notable milestones in the understanding of ai over the past decade, especially with the rapid rise of deep learning research. However, much of the ocean remains unexplored and the coast still seems far away for real-world applications. At every moment, researchers from different parts of the world aspire and strive to innovate ai solutions that are implementable in real and challenging applications. In this article we talk about the Thousand Brains Project. This ongoing research project aims to develop a complementary form of ai based on the working principles of the human brain, specifically the neocortex. Imagine your machine spends a few glances at an object, and the next second it knows how to grab and interact with the object via touch for final execution. This is something that Thousand Brains Project is working. TO team of researchers from Numenta Inc. aims to develop a universal platform for flexible and capable robotics and artificial intelligence solutions that differ from deep learning approaches. Additionally, this project allows models to seamlessly use inputs from multiple modalities.

This project differs from contemporary technologies because it uses sensorimotor learning at its core. Sensorimotor interactive systems learn by detecting different parts of the world while constantly interacting; Here, learnings represent the foundational knowledge that supports other functions. Additionally, the researchers have chosen a design that differentiates sensory and motor processing with separate architectures. They choose actively generated temporal sequences of sensory and motor inputs to achieve this. Three components constitute the architecture: sensor modules, learning modules and the motor system. A final key component, a common communication protocol, unites these three elements.

This project uses explicit reference frames and coordinate systems in their learned structured models, providing a second order of differentiation. Reference frames track the positions of sensors relative to the environment and are used at all levels of information processing, including the environment, objects, and even abstract ideas. These models follow a policy of “knowing” and “growing”, as they constantly learn from their environment and manipulate objects accordingly. based on estimates from structured data.

He The Thousand Brains Project is quite ambitious and strives to build a universal platform and messaging protocol for intelligent sensorimotor systems in the long term. To do this, the team is developing an interface called Cortical Messaging Protocol. CMP enables rapid project scaling and helps with intermodular communication over the defined channel.

Now we will talk about the first implementation of TBP: Monty. Monty has sensor modules to convert raw sensory data into a common language, learning modules to model incoming data streams and use them for interaction, and a motor system for execution. CMP also links all its components. Currently, the authors have focused on object recognition and pose detection tasks. For the experiment, the agents interacted with the environment and collected observations. The authors describe Monty's differentiated phase of training and evaluation. The article then mentions several Monty apps in multiple configurations.

Conclusion: The Thousand Brains Project is developing a new field of artificial intelligence inspired by the human brain. This ambitious project aims to present a holistic artificial intelligence and robotics solution. The project's models are currently working on object recognition tasks; However, this is broad in scope and some potential future tasks could be language modeling, learning without explicit supervision, generalizing modeling, etc.

Verify he Paper and GitHub page. All credit for this research goes to the researchers of this project. Also, don't forget to follow us on <a target="_blank" href="https://twitter.com/Marktechpost”>twitter and join our Telegram channel and LinkedIn Grabove. Don't forget to join our SubReddit over 60,000 ml.

UPCOMING FREE ai WEBINAR (JANUARY 15, 2025): <a target="_blank" href="https://info.gretel.ai/boost-llm-accuracy-with-sd-and-evaluation-intelligence?utm_source=marktechpost&utm_medium=newsletter&utm_campaign=202501_gretel_galileo_webinar”>Increase LLM Accuracy with Synthetic Data and Assessment Intelligence–<a target="_blank" href="https://info.gretel.ai/boost-llm-accuracy-with-sd-and-evaluation-intelligence?utm_source=marktechpost&utm_medium=newsletter&utm_campaign=202501_gretel_galileo_webinar”>Join this webinar to learn practical information to improve LLM model performance and accuracy while protecting data privacy..