Researchers say AI models like GPT4 are prone to “sudden” escalations as the U.S. military explores their use for warfare.
- Researchers ran international conflict simulations with five different AIs and found that they tended to escalate war, sometimes out of nowhere, and even use nuclear weapons.
- The AIs were large language models (LLMs) like GPT-4, GPT 3.5, Claude 2.0, Llama-2-Chat, and GPT-4-Base, which are being explored by the U.S. military and defense contractors for decision-making.
- The researchers invented fake countries with different military levels, concerns, and histories and asked the AIs to act as their leaders.
- The AIs showed signs of sudden and hard-to-predict escalations, arms-race dynamics, and worrying justifications for violent actions.
- The study casts doubt on the rush to deploy LLMs in the military and diplomatic domains, and calls for more research on their risks and limitations.
LLMs are an attempt to develop artificial intelligence essentially through “simple complex systems”. The argument being that’s how human intelligence is essentially work.
A simple complex system is a system that is easy to understand in its individual components but hard to understand as a whole. Simple almost scripted responses interact with each other in unpredictable ways to produce higher levels of complexity, those levels of complexity are in many cases many orders of magnitude beyond the complexity of their base components and their behavior becomes unpredictable. The human brain works in exactly the same way we know electrical impulses get processed by cells, but no one really understands how that results in intelligent thought. Sounds like an AI to me.