Decisions are constantly made in our daily lives. Whether it is to restock the morning cereal in the canister or watch a newly released Netflix series late at night, our decisions are more or less indicative of how we put value into the actions we make. Therefore, our actions often reflect how we try to maximize our benefits in life by making what we believe are efficient choices. To make the best decision from a range of options, we may need to think systematically, a skill we usually develop through education. Optimization refers to improving efficiency and productivity through our decisions, but this ability is not unique to humans. By observing how animals optimize their actions, we can better understand the roots of our own decision-making processes.

Even monkeys, it turns out, have a knack for this! In an experiment by researchers at Carnegie Mellon University and the University of Pittsburgh, two rhesus macaque monkeys were given a virtual backpack and asked to decide which items to ‘pack’ for the best rewards [1]. Through a combinatorial optimization task, the researchers were able to determine the performance of the monkeys in a knapsack task where the monkeys had to identify valuable items and create combinations to fit optimally into the knapsack. The monkeys were rewarded with juice when they correctly recognized the fractal images representing the items. To measure how well the monkeys chose items that would maximize their rewards, the researchers used the Houtman–Maks (HM) index. The index indicated how close the monkeys' choices were to being the best possible, a measure of how effectively they could make decisions [1].

To further explore their decision-making, the researchers gave monkeys two types of tasks with different levels of difficulty: first a simpler task of identifying the value of an item, and second a more complex task of optimizing their choices [1]. To better understand the connection between the monkeys' decision-making and these tasks, the researchers used a recurrent neural network. This model helped reveal how the monkeys' rational needs and thought processes aligned with patterns like clustering and thresholding, which guided their final choices [1].

Despite the insights provided by the combinatorial strategies to analyze the optimization in decision-making, one fascinating question remains: What truly drives algorithmic selection in decision-making? Many of our choices—whether it’s choosing what to eat or deciding how to spend our time—involve careful thought and mental deliberation. Just like the rhesus monkeys in the study, our brains work to balance different factors, using systematic thinking to maximize our rewards and minimize costs. The study's findings reveal that even non-human primates use complex decision-making processes, challenging the idea that optimization is a uniquely human trait. By examining how the monkeys tackled tasks of varying difficulty, researchers have laid the groundwork for a deeper understanding of algorithmic-based decision-making and economic deliberation. Future studies could build on these insights to explore how the brain handles complex choices, offering computational models to investigate the neural basis of decision-making in humans.

Reference: 

[1] Hong, T., & Stauffer, W. R. (2023). Computational complexity drives sustained deliberation. Nature neuroscience, 26(5), 850–857. https://doi.org/10.1038/s41593-023-01307-6