Chemical reactions are often described using simple equations that show starting materials turning into products. However, in reality, the same substances can react in many different ways depending on factors like concentration, temperature, and the presence of catalysts. The sheer number of possible reaction pathways makes it nearly impossible for scientists to explore them all manually. Recent advancements in laboratory automation and chemical artificial intelligence are helping researchers systematically explore this vast "reaction hyperspace," uncovering new chemical processes that may have remained hidden even in long-studied reactions. A team of scientists led by Bartosz A. Grzybowski, director of the Center for Algorithmic and Robotized Synthesis at the Institute for Basic Science (IBS), used an automated robotic system to investigate 960 different conditions for the Biginelli reaction, a classic chemical process first reported in 1891. Rather than searching for the best way to produce a known molecule, the researchers aimed to discover all the possible products and reaction pathways that could arise under various conditions. Their work, published in Nature Synthesis, revealed a previously unknown variation of the Biginelli reaction that forms complex, bicyclic structures not typically seen in this reaction. Further analysis, supported by chemical AI, showed that this new pathway involves a pseudo-seven-component transformation, where seven starting molecules contribute to the formation of a single complex product. Using this newly discovered reaction network, the researchers redesigned the synthesis and created a family of related molecules. Some of these molecules displayed structural complexity similar to natural products, which are often highly complex and difficult to synthesize. The newly discovered molecules were also notable for their unique behaviors at the supramolecular level. Some compounds spontaneously assembled into larger structures depending on their concentration and temperature. Others selectively bound specific metal ions, such as barium and zinc, suggesting possible applications in metal sensing. One compound exhibited a rare form of chiral self-sorting, where molecules exist as mirror-image forms known as enantiomers. Depending on the environment—such as solid state versus solution or the presence of specific metal ions—the compound favored assembling with molecules of the same or opposite handedness. This kind of metal-programmable chiral sorting is extremely rare and could have applications in areas like sensing, responsive materials, and molecular recognition. The broader significance of this research lies in the method used to discover the reaction. Traditional automated chemistry typically focuses on optimizing the production of a known product. In this case, the robotic platform was used to explore the full reaction network, revealing unexpected products and pathways in regions of chemical space that are usually overlooked. The researchers suggest that this "hyperspace" approach could shift the role of chemical automation from simply speeding up experiments to enabling the discovery of entirely new chemical processes. Even reactions studied for over a century may still hold hidden pathways that become apparent only when explored systematically. This study shows that combining robotic experimentation, large-scale reaction mapping, and chemical AI can uncover new reaction mechanisms and complex molecules with surprising functional properties.