How do we predict others from patterns in their behavior and what are the computational constraints that limit this ability? We investigate these questions by modeling human behavior over repeated games of rock, paper, scissors from Brockbank & Vul (2024). Against algorithmic opponents that varied in strategic sophistication, people readily exploit simple transition patterns (e.g., consistently playing rock after paper) but struggle to detect more complex sequential dependencies. To understand the cognitive mechanisms underlying these abilities and their limitations, we deploy Hypothetical Minds (HM), a large language model-based agent that generates and tests hypotheses about opponent strategies, as a cognitive model of this behavior (Cross et al., 2024). We show that when applied to the same experimental conditions, HM closely mirrors human performance patterns, succeeding and failing in similar ways. To better understand the source of HM's failures and whether people might face similar cognitive bottlenecks in this context, we performed a series of ablations and augmentations targeting different components of the system. When provided with natural language descriptions of the opponents' strategies, HM successfully exploited 6/7 bot opponents with win rates >80% suggesting that accurate hypothesis generation is the primary cognitive bottleneck in this task. Further, by systematically manipulating the model's hypotheses through pedagogically-inspired interventions, we find that the model substantially updates its causal understanding of opponent behavior, revealing how model-based analyses can produce testable hypotheses about human cognition.

Game theory, as an analytical tool, is frequently utilized to analyze human behavior in social science research. With the high alignment between the behavior of Large Language Models (LLMs) and humans, a promising research direction is to employ LLMs as substitutes for humans in game experiments, enabling social science research. However, despite numerous empirical researches on the combination of LLMs and game theory, the capability boundaries of LLMs in game theory remain unclear. In this research, we endeavor to systematically analyze LLMs in the context of game theory. Specifically, rationality, as the fundamental principle of game theory, serves as the metric for evaluating players' behavior -- building a clear desire, refining belief about uncertainty, and taking optimal actions. Accordingly, we select three classical games (dictator game, Rock-Paper-Scissors, and ring-network game) to analyze to what extent LLMs can achieve rationality in these three aspects. The experimental results indicate that even the current state-of-the-art LLM (GPT-4) exhibits substantial disparities compared to humans in game theory. For instance, LLMs struggle to build desires based on uncommon preferences, fail to refine belief from many simple patterns, and may overlook or modify refined belief when taking actions. Therefore, we consider that introducing LLMs into game experiments in the field of social science should be approached with greater caution.

Progress in fields of machine learning and adversarial planning has benefited significantly from benchmark domains, from checkers and the classic UCI data sets to Go and Diplomacy. In sequential decision-making, agent evaluation has largely been restricted to few interactions against experts, with the aim to reach some desired level of performance (e.g. beating a human professional player). We propose a benchmark for multiagent learning based on repeated play of the simple game Rock, Paper, Scissors along with a population of forty-three tournament entries, some of which are intentionally sub-optimal. We describe metrics to measure the quality of agents based both on average returns and exploitability. We then show that several RL, online learning, and language model approaches can learn good counter-strategies and generalize well, but ultimately lose to the top-performing bots, creating an opportunity for research in multiagent learning.

Recent multi-agent extensions of Q-Learning require knowledge of other agents' payoffs and Q-functions, and assume game-theoretic play at all times by all other agents. This paper proposes a fundamentally different approach, dubbed "Hyper-Q" Learning, in which values of mixed strategies rather than base actions are learned, and in which other agents' strategies are estimated from observed actions via Bayesian in- ference. Hyper-Q may be effective against many different types of adap- tive agents, even if they are persistently dynamic. Against certain broad categories of adaptation, it is argued that Hyper-Q may converge to ex- act optimal time-varying policies. In tests using Rock-Paper-Scissors, Hyper-Q learns to significantly exploit an Infinitesimal Gradient Ascent (IGA) player, as well as a Policy Hill Climber (PHC) player.

Recently, "AI" machine-learning technologies have been creeping their way into artistic fields in both entertaining and harmful ways. While some AI content creators are just making videos for harmless fun, others, like the creators of a recent AI-generated anime short, wrongfully believe they've democratized the animation industry when they've really just come up with a more technologically demanding method of plagiarizing other artists. Earlier this week, Corridor Digital, a Los Angeles-based production studio that creates pop culture YouTube videos, uploaded a video called "Anime Rock, Paper, Scissors." Written and directed by Niko Pueringer and Sam Gorski, it revolves around two twins vying for the throne left vacant by their recently deceased father. By leveraging the machine-learning text-to-image model Stable Diffusion, Corridor Digital gave camera footage filmed in front of a green screen a dramatic anime-like appearance.

Pricing decisions are increasingly made by AI. Thanks to their ability to train with live market data while making decisions on the fly, deep reinforcement learning algorithms are especially effective in taking such pricing decisions. In e-commerce scenarios, multiple reinforcement learning agents can set prices based on their competitor's prices. Therefore, research states that agents might end up in a state of collusion in the long run. To further analyze this issue, we build a scenario that is based on a modified version of a prisoner's dilemma where three agents play the game of rock paper scissors. Our results indicate that the action selection can be dissected into specific stages, establishing the possibility to develop collusion prevention systems that are able to recognize situations which might lead to a collusion between competitors. We furthermore provide evidence for a situation where agents are capable of performing a tacit cooperation strategy without being explicitly trained to do so.

In this article, I'll walk you through my process of building a full stack python Flask artificial intelligence project capable of beating the human user over 60% of the time using a custom scoring system to ensemble six models (naïve logic-based, decision tree, neural network) trained on both game-level and stored historical data in AWS RDS Cloud SQL database. Rock Paper Scissors caught my attention for an AI project because, on the surface, it seems impossible to get an edge in the game. These days, it is easy to assume that a computer can beat you in chess, because it can harness all of its computing power to see all possible outcomes and choose the ones that benefit it. Rock Paper Scissors, on the other hand, is commonly used in place of a coin toss to solve disputes because the winner seems random. My theory though, was that humans can't actually make random decisions, and that if an AI could learn to understand the ways in which humans make their choices over the course of a series of matches, even if the human was trying to behave randomly, then the AI would be able to significantly exceed 33% accuracy in guessing the player's decisions.

In this article, I'll walk you through my process of building a full stack python Flask artificial intelligence project capable of beating the human user over 60% of the time using a custom scoring system to ensemble six models (naïve logic-based, decision tree, neural network) trained on both game-level and stored historical data in AWS RDS Cloud SQL database. Rock Paper Scissors caught my attention for an AI project because, on the surface, it seems impossible to get an edge in the game. These days, it is easy to assume that a computer can beat you in chess, because it can harness all of its computing power to see all possible outcomes and choose the ones that benefit it. Rock Paper Scissors, on the other hand, is commonly used in place of a coin toss to solve disputes because the winner seems random. My theory though, was that humans can't actually make random decisions, and that if an AI could learn to understand the ways in which humans make their choices over the course of a series of matches, even if the human was trying to behave randomly, then the AI would be able to significantly exceed 33% accuracy in guessing the player's decisions.

Predicting and modeling human behavior and finding trends within human decision-making process is a major social sciences'problem. Rock Paper Scissors (RPS) is the fundamental strategic question in many game theory problems and real-world competitions. Finding the right approach to beat a particular human opponent is challenging. Here we use Markov Chains with set chain lengths as the single AIs (artificial intelligences) to compete against humans in iterated RPS game. This is the first time that an AI algorithm is applied in RPS human competition behavior studies. We developed an architecture of multi-AI with changeable parameters to adapt to different competition strategies. We introduce a parameter "focus length" (an integer of e.g. 5 or 10) to control the speed and sensitivity for our multi-AI to adapt to the opponent's strategy change. We experimented with 52 different people, each playing 300 rounds continuously against one specific multi-AI model, and demonstrated that our strategy could win over more than 95% of human opponents.

In the first part, we explored how Bayesian Statistics might be used to make reinforcement learning less data-hungry. Now we execute this idea in a simple example, using Tensorflow Probability to implement our model. When it comes to games, it is difficult to imagine something simpler than rock, paper, scissors. Despite the simplicity, googling the game reveals a remarkable body of literature. We want to use Bayesian Statistics to play this game and exploit the biases of a human opponent.