The Thousand Brains Theory: Applying Neocortex Principles to Decision-Making & Business Strategy
"Empires of the future will be empires of the mind." - Winston Churchill
Last week, we explored how human intelligence evolved. This week, we're diving into the groundbreaking work of Vernon Mountcastle and Jeff Hawkins to uncover how our brains make decisions. By applying their insights, we can create more intelligent business systems that make better decisions and ultimately drive business success.
Vernon Mountcastle’s Cortical Columns
Vernon Mountcastle discovered that the neocortex (the part of the brain responsible for higher-order “executive” functions) comprises thousands of tiny columns of neurons, called "cortical columns." Each column processes specific types of sensory information, like touch or vision. Mountcastle suggested that these columns work similarly across different parts of the brain. Each column operates within a larger network, contributing to the brain's overall functionality.
Expanding on Mountcastle: Jeff Hawkins "Thousand Brains Theory"
Jeff Hawkins took Mountcastle's idea further with his "Thousand Brains Theory." He proposed that each cortical column can learn complete models of objects or concepts based on sensory input, which he calls "reference frames." Take an apple for example, one cortical column makes a model of the apple's colour, another makes a model of its shape, another of its smell, and so on. Each column has its complete picture or "reference frame" of the apple. Hawkins believes intelligence comes from combining the outputs of many cortical columns, each acting like a mini-brain, to create a complete understanding of the world.
Coordinated Collective Intelligence
Mountcastle's and Hawkins' findings suggest that the brain's "brain of brains" system integrates outputs from distributed cortical columns to make accurate decisions. Each column has a specialized role, and their collective outputs are coordinated through a hierarchical system, synthesizing inputs, prioritizing relevant information, and guiding the right decision through collective intelligence.
"Coordinated Collective Intelligence" System - A Simple Business Analogy
Let's break it down into a simple analogy to understand how this system works.
Independent Departments: Each department in your company (like marketing, finance, and R&D) is independent. Each one learns and processes information relevant to its area, similar to how cortical columns in the brain learn about different aspects of sensory inputs.
Central Executive Team: Now, think of a central executive team that brings together insights from all these departments. This team integrates the outputs from each department to form a complete picture of the company’s situation.
Receiving Outputs from Departments: Each department sends its predictions, insights, and data to the executive team in parallel, just like cortical columns send their learned models to the higher brain.
Combining Models: The executive team combines these diverse insights, to create a comprehensive understanding of the business environment.
Extracting Critical Information: The executive team identifies the most critical information from these combined insights, focusing on what’s most relevant for strategic decision
Prioritizing and Combining: By prioritizing this essential information, the executive team decides on the best course of action that aligns with the overall business goals.
Higher-Level Strategic Decisions: This hierarchical structure enables the company to make well-informed, strategic decisions that couldn’t be achieved by any single department alone. It leverages the combined knowledge of all departments to predict outcomes and adapt effectively.
Simplified Framework for Balancing Multiple Objectives with AI
One of the hardest parts of this in the real world is defining the business goal (or the objective you are trying to achieve). For humans, our aim is to maximize survival, safety, and well-being. However, it is slightly more complicated in the corporate world since we must balance multiple objectives simultaneously.
Balancing multiple objectives is challenging given the limitations of today's narrow AI algorithms, which typically excel in optimizing for a single, well-defined objective function. However, I hypothesize by structuring your approach using a hierarchical system, you can achieve effective decision-making and optimization. Here’s a simplified framework:
1. Hierarchical Objective Function Framework
Primary Objective: Define an overarching goal that aligns with your company’s mission (e.g., Increase Revenue).
Secondary Objectives: Break down into specific, actionable objectives (i.e. increase acquisition, customer satisfaction, optimize resource output, minimize risks, etc.)
2. Specialized AI Models for Secondary Objectives
Develop AI models focused on each secondary objective (some examples below):
Acquisition Optimization: Analyze sales data, customer behaviour, growth opportunities, and market trends to maximize acquisition
Resource Utilization: Optimize resource allocation, inventory management, and workforce planning
Customer Satisfaction: Analyze customer feedback and interactions to enhance customer experience and loyalty.
Risk Management: Anomaly detection or activity beyond expectations to identify potential risks
3. Integration and Prioritization by a Central System
Central System: Performs the executive function integrating outputs from specialized models and data, includes:
Data Aggregation: Collect data from all models in real-time to get a comprehensive view.
Weighting Mechanism: Prioritize objectives based on their importance and current business context (e.g., prioritize risk management during a market downturn).
Multi-Objective Optimization: Use techniques designed to simultaneously optimize two or more conflicting objectives subject to certain constraints.
Scenario Analysis and Simulation: Evaluate the impact of different decisions on multiple objectives to make informed trade-offs and adjustments.
4. Continuous Learning and Adaptation
Feedback Loops: Continuously monitor the performance of the system’s decisions and your primary and secondary metrics
Adaptive Algorithms: Adjust weights and models based on new data and changing business conditions to improve accuracy and effectiveness.
Conclusion
As we keep exploring the amazing possibilities of AI and machine learning, it's essential to consider how such a system can transform how we run our businesses and make decisions. By using this framework and digitizing our processes, we can revolutionize decision-making & business strategy, making companies more flexible, efficient, and profitable. Next week, we'll dive into a real-world example of how a company can put this framework into action and where human expertise is crucial to making it all work.
Call to Action:
Consider how you can structure an AI system to balance multiple objectives and enhance your business’s strategic decision-making. Share your thoughts and experiences in the comments below!