Marcin Korecki

Marcin Korecki, Damian Dailisan, Cesare Carissimo · eth-zurich

The development of ethical AI systems is currently geared toward setting objective functions that align with human objectives. However, finding such functions remains a research challenge, while in RL, setting rewards by hand is a fairly standard approach. We present a methodology for dynamic value alignment, where the values that are to be aligned with are dynamically changing, using a multiple-objective approach. We apply this approach to extend Deep $Q$-Learning to accommodate multiple objectives and evaluate this method on a simplified two-leg intersection controlled by a switching agent.Our approach dynamically accommodates the preferences of drivers on the system and achieves better overall performance across three metrics (speeds, stops, and waits) while integrating objectives that have competing or conflicting actions.

Joshua C. Yang, Damian Dailisan, Marcin Korecki et al. · eth-zurich

This paper investigates the voting behaviors of Large Language Models (LLMs), specifically GPT-4 and LLaMA-2, their biases, and how they align with human voting patterns. Our methodology involved using a dataset from a human voting experiment to establish a baseline for human preferences and conducting a corresponding experiment with LLM agents. We observed that the choice of voting methods and the presentation order influenced LLM voting outcomes. We found that varying the persona can reduce some of these biases and enhance alignment with human choices. While the Chain-of-Thought approach did not improve prediction accuracy, it has potential for AI explainability in the voting process. We also identified a trade-off between preference diversity and alignment accuracy in LLMs, influenced by different temperature settings. Our findings indicate that LLMs may lead to less diverse collective outcomes and biased assumptions when used in voting scenarios, emphasizing the need for cautious integration of LLMs into democratic processes.

Marcin Korecki, Cesare Carissimo

This paper investigates the concept of Labour as an expression of `timenergy' - a fusion of time and energy - and its entanglement within the system of Capital. We define Labour as the commodified, quantifiable expansion of timenergy, in contrast to Capital, which is capable of accumulation and abstraction. We explore Labour's historical evolution, its coercive and alienating nature, and its transformation through automation and artificial intelligence. Using a game-theoretic, agent-based simulation, we model interactions between Capital and Labour in production processes governed by Cobb-Douglas functions. Our results show that despite theoretical symmetry, learning agents disproportionately gravitate toward capital-intensive processes, revealing Capital's superior organizational influence due to its accumulative capacity. We argue that Capital functions as an artificially alive system animated by the living Labour it consumes, and question whether life can sustain itself without the infrastructures of Capital in a future of increasing automation. This study offers both a critique of and a framework for understanding Labour's subjugation within the Capital system.

Marcin Korecki

The dominant paradigm in AI ethics and value alignment is highly anthropocentric. The focus of these disciplines is strictly on human values which limits the depth and breadth of their insights. Recently, attempts to expand to a sentientist perspective have been initiated. We argue that neither of these outlooks is sufficient to capture the actual complexity of the biosphere and ensure that AI does not damage it. Thus, we propose a new paradigm -- Biospheric AI that assumes an ecocentric perspective. We discuss hypothetical ways in which such an AI might be designed. Moreover, we give directions for research and application of the modern AI models that would be consistent with the biospheric interests. All in all, this work attempts to take first steps towards a comprehensive program of research that focuses on the interactions between AI and the biosphere.