Thinking in maps: from the Lascaux caves to knowledge graphs

Thinking in maps: from the Lascaux caves to knowledge graphs

The concept of thinking in maps has been ingrained in human history for millennia. When we think of maps, we often associate them with geography, but maps can be any symbolic representation of the relationship between elements of a physical or mental space. From the Lascaux caves to knowledge graphs, maps have played a crucial role in how we understand and communicate information.

In the Lascaux caves, one of the oldest known maps in the world can be found. Interestingly, it is not a map of the earth, but a map of the skies. This shows that maps have been used to depict various aspects of our world, not just geographical locations. Maps were not only used as communication tools but also as a means to express religious and artistic aspects. The hieroglyphs in ancient Egypt, for example, served as a visual representation that went beyond simple communication.

Throughout history, various cultures have used maps in different forms. Pictograms and symbols are common elements found in almost all cultures, used to convey meaning alongside written language. Even road signs can be seen as mini-maps, using symbolic shapes and colors to connect ideas and create new meaning. Leonardo da Vinci, known for his polymathic nature, saw the interconnectedness of art and science, stating that "everything connects to everything else." Isaac Newton also employed diagrams and sketches to explore scientific concepts and ideas.

While the term "mind map" was coined in 1974 by Tony Buzan, the practice of thinking in maps has been present throughout history. Different types of maps have been developed to facilitate this process. Radial maps, such as traditional mind maps, start from a central concept and expand outward. Nested maps, on the other hand, start from a general concept and grow inward. Topic maps allow for interlinking between topics without following a radial structure, while process maps add directionality by connecting concepts in a specific order. Concept maps, developed by Professor Joseph D. Novak, use a context frame to enhance meaningful learning.

One of the challenges of thinking in maps is managing the complexity and messiness of our thought processes. This is where knowledge graphs come into play. Google's Knowledge Graph, for example, has connected billions of concepts with minimal human intervention. Knowledge graphs represent information in a formalized way, describing concepts, categories, and relationships. They have become one of the most efficient methods for visualizing ontologies. The memex, a concept that influenced the development of hypertext systems, paved the way for the creation of the Internet as we know it today.

In the realm of knowledge management, tools like Roam and Notion are exploring different approaches. Roam offers automated bi-directional linking and a visual knowledge graph, while Notion databases take a hierarchical approach to information. The next step in this journey is the metamodeling of thinking in maps, where a new language and schema can be developed to represent our thought processes. This metamodel holds the potential for a revolution in collective intelligence.

Moving away from the realm of maps, we enter the world of generative AI and the future of search engines. The current search engine paradigm is rooted in the technology of the late 1990s, and while it has served us well, it may not be the best way to search in today's content landscape. The way we consume information has evolved, with a significant amount of content being presented in graph form, data streams, video content, and more.

The idea of a generative AI search engine challenges the traditional approach. Instead of searching a database and scanning through results, a generative AI search engine would use the database as training data and generate the desired answer. Trained models, although much smaller in size compared to the training data, have the potential to provide more accurate and relevant results.

This shift in search engine design would bypass the distribution monopoly and advertising business of incumbents. Training a model may be expensive, but running it has minimal marginal costs. This poses interesting questions about the future of advertising in this new paradigm.

In conclusion, the concept of thinking in maps has been ingrained in human history, from ancient cave paintings to modern knowledge graphs. Maps have been used as a means of communication, expression, and understanding. The future of thinking in maps lies in the development of metamodels and the utilization of generative AI in search engines. As we continue to explore new ways of organizing and accessing information, the potential for collective intelligence and revolutionary advancements in technology is within reach.

Three actionable advice before conclusion:

1. Embrace the interconnectedness of knowledge: Like polymaths such as Leonardo da Vinci, strive to see the links between different disciplines and areas of knowledge. This holistic approach can lead to new insights and creative solutions.

2. Explore different mapping techniques: Experiment with various mapping techniques, such as radial maps, nested maps, topic maps, process maps, and concept maps. Each approach offers unique ways to organize and visualize information.

3. Stay open to technological advancements: Keep an eye on emerging technologies like generative AI and knowledge graphs. These innovations have the potential to reshape how we think, learn, and search for information. Stay curious and embrace the possibilities they offer.

In the quest for collective intelligence, the future lies in thinking in maps and leveraging the power of generative AI. As we continue to push the boundaries of knowledge and technology, the tools we use will eventually catch up with the way our minds naturally think, enabling a new era of interconnectedness and understanding.