Photo by Editor
# Understanding Just-in-Time World Modeling
This article provides an overview and summary of a recently published paper titled “Just in Time” Earth Modeling Supports Human Resource Planning, which is available in full to read at. arXiv.
Using a softer and more accessible tone for a wider audience, we will introduce what is simulation-based thinking, explain the general just-in-time (JIT) framework presented in the topic with a focus on the planning of the methods it uses, and summarize how it behaves and helps to develop assumptions in the context of supporting human planning and thinking.
# Understanding Simulation-Based Reasoning
Imagine you are in the far corner of a dark, dirty room full of obstacles and you want to determine the correct way to reach the door without colliding. In parallel, suppose you are about to hit a pool ball and visualize the path you expect the ball to follow. In these two situations, there is one thing in common: the ability to project a future situation in our mind without doing any action. This is known as simulation-based reasoningand sophisticated AI agents that require this ability in a variety of situations.
Simulation-based thinking is a cognitive tool that we humans constantly use to make decisions, plan routes, and predict what will happen next in our environment. However, the real world is incredibly complex and full of nuances and details. Trying to fully account for all possible outcomes and their implications can quickly exhaust our mental resources in a matter of milliseconds. To avoid this, in biological terms, what we are doing is not creating an almost perfect copy of the image in reality, but rather producing a simplified representation that retains only the truly relevant information.
The scientific community is still trying to answer a big question: How does our brain decide quickly and efficiently which information to include and which to skip in that mental simulation? That question motivates the JIT framework presented in the target study.
# Assessing Basic Methods
In order to answer the previously formulated question, the researchers in the study present a new JIT framework that, unlike traditional theories that take full environmental observation before planning, proposes to build a mental map on the fly, collecting information only when it is really needed.
The JIT framework proposed in the paper and applied to the navigation problem | Source: here
The main success of this model is how it describes the combination and interplay between three important methods:
- Imitation: It is based on the principle that our mind starts planning in advance the course of action or the route we will follow.
- Visual search: As the mental simulation moves into the unknown, it sends our eyes (or visions, in the case of AI agents or systems) a signal to explore that particular part of the physical (or digital) environment.
- Transformation of representation: When an object that may interfere with our system is detected, eg an obstacle, the brain immediately “combines” that object and adds it to its mental model for consideration.
In essence, this is a fast and smooth cycle: The brain simulates at a low level, then the “eyes” search for obstacles, the mind updates the information, and the simulation continues – all in a well-organized way.
# Organizational Behavior and Its Impact on Decision Making
What is the most interesting aspect of the JIT model presented in the paper? It is disputed which works amazingly well. The authors tested it by comparing human behavior with computer simulations in two experiments: navigating a maze and physical prediction tests, such as predicting where a ball will bounce.
The results showed that a JIT system keeps in memory a much smaller number of objects than systems that try to process the entire space from scratch. However, despite operating based on a fragmented mental picture that includes only a small part of the full reality, the framework is able to make high-level, informed decisions. This gives a deep take: Our brain improves its performance and speed of response not by processing more data, but by being more selective, achieving reliable predictions without spending too much cognitive effort.
# Considering Future Directions
Although the JIT framework presented in the study provides a clear explanation of how humans plan (with potential implications for pushing the boundaries of AI systems), there are some issues that remain to be addressed. The survey conducted in the study only looked at static areas. Therefore, extending this model should also consider dynamic and chaotic conditions. Understanding how important information is selected when many constants coexist around us may be the next big challenge to drive progress in this fascinating human programming theory and theory of mind and — who knows! – translating it into the world of AI.
Iván Palomares Carrascosa is a leader, author, speaker, and consultant in AI, machine learning, deep learning and LLMs. He trains and guides others in using AI in the real world.
