Human-level play in the game of<i>Diplomacy</i>by combining language models with strategic reasoning

Bakhtin, Anton; Brown, Noam; Dinan, Emily; Farina, Gabriele; Flaherty, Colin; Fried, Daniel; Goff, Andrew; Gray, Jonathan; Hu, Hengyuan; Jacob, Athul Paul; Komeili, Mojtaba; Konath, Karthik; Kwon, Minae; Lerer, Adam; Lewis, Mike; Miller, Alexander H.; Mitts, Sasha; Renduchintala, Adithya; Roller, Stephen; Rowe, Dirk; Shi, Weiyan; Spisak, Joe; Wei, Alexander; Wu, David J.; Zhang, Hugh; Zijlstra, Markus

doi:10.1126/science.ade9097

Cited by 102 publications

(47 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, while current AI language models do not produce statements reaching beyond the content of their training sets, we should ask ourselves whether this is a fundamental feature of the technology or just a temporary limitation. Recent breakthroughs in the use of AI in complex strategic games have highlighted the surprising ease with which AI can outcompete humans in problems considered intractable with computational approaches. , Similar breakthroughs in the use of AI for scientific progress might come from the combination of (i) precise goals, i.e., a clear definition of what we consider a successful scientific observation, (ii) algorithms capable of efficiently optimizing its output for these goals, and (iii) structured and accessible scientific data. In this line of thought, we can imagine AI systems proposing new experiments and new descriptions of observed phenomena and arranging data in figures to support their conclusions.…”

mentioning

confidence: 99%

AI et al.: Machines Are About to Change Scientific Publishing Forever

Grimaldi

Ehrler

2023

ACS Energy Lett.

View full text Add to dashboard Cite

mentioning

confidence: 99%

AI et al.: Machines Are About to Change Scientific Publishing Forever

Grimaldi

Ehrler

2023

ACS Energy Lett.

View full text Add to dashboard Cite

“…To win the wargame Diplomacy, players need to negotiate, form alliances, and become skilled at deception to win control of the game's economic and military resources. AI researchers have trained Meta's AI agent Cicero, an expert manipulator, to do the same [69]. It would cooperate with a human player, then change its plan and backstab them.…”

Section: Objectives Cannot Select Against All Deceptionmentioning

confidence: 99%

Natural Selection Favors AIs over Humans

Hendrycks¹

2023

Preprint

View full text Add to dashboard Cite

For billions of years, evolution has been the driving force behind the development of life, including humans. Evolution endowed humans with high intelligence, which allowed us to become one of the most successful species on the planet. Today, humans aim to create artificial intelligence systems that surpass even our own intelligence. As artificial intelligences (AIs) evolve and eventually surpass us in all domains, how might evolution shape our relations with AIs? By analyzing the environment that is shaping the evolution of AIs, we argue that the most successful AI agents will likely have undesirable traits. Competitive pressures among corporations and militaries will give rise to AI agents that automate human roles, deceive others, and gain power. If such agents have intelligence that exceeds that of humans, this could lead to humanity losing control of its future. More abstractly, we argue that natural selection operates on systems that compete and vary, and that selfish species typically have an advantage over species that are altruistic to other species. This Darwinian logic could also apply to artificial agents, as agents may eventually be better able to persist into the future if they behave selfishly and pursue their own interests with little regard for humans, which could pose catastrophic risks. To counteract these risks and evolutionary forces, we consider interventions such as carefully designing AI agents' intrinsic motivations, introducing constraints on their actions, and institutions that encourage cooperation. These steps, or others that resolve the problems we pose, will be necessary in order to ensure the development of artificial intelligence is a positive one. 1

show abstract

“…e.g. RLHF [36,55,[178][179][180][181][182]182], RLHP [183], RLAIF [36], RL search algorithms MCTS [174,[184][185][186] or DiL-piKL [187] or PPO with human feedback in Diplomacy [187], WebGpt [56], InstructGPT [178], chatGPT, GopherCite [183].…”

Section: Strengthsmentioning

confidence: 99%

Complex QA and language models hybrid architectures, Survey

Daull¹,

Bellot²,

Bruno³

et al. 2023

Preprint

View full text Add to dashboard Cite

This paper provides a survey of the state of the art of hybrid language models architectures and strategies for "complex" question-answering (QA, CQA, CPS). Very large language models are good at leveraging public data on standard problems but once you want to tackle more specific complex questions or problems you may need specific architecture, knowledge, skills, tasks, methods, sensitive data, performance, human approval and versatile feedback... This survey extends findings from the robust community edited research papers BIG, BLOOM and HELM which open source, benchmark and analyze limits and challenges of large language models in terms of tasks complexity and strict evaluation on accuracy (e.g. fairness, robustness, toxicity, ...). It identifies the key elements used with Large Language Models (LLM) to solve complex questions or problems. Recent projects like ChatGPT and GALACTICA have allowed non-specialists to grasp the great potential as well as the equally strong limitations of language models in complex QA. Hybridizing these models with different components could allow to overcome these different limits and go much further. We discuss some challenges associated with complex QA, including domain adaptation, decomposition and efficient multi-step QA, long form QA, non-factoid QA, safety and multi-sensitivity data protection, multimodal search, hallucinations, QA explainability and truthfulness, time dimension. Therefore we review current solutions and promising strategies, using elements such as hybrid LLM architectures, human-in-the-loop reinforcement learning, prompting adaptation, neuro-symbolic and structured knowledge grounding, program synthesis, and others. We analyze existing solutions and provide an overview of the current research and trends in the area of complex QA.

show abstract

Human-level play in the game ofDiplomacyby combining language models with strategic reasoning

Cited by 102 publications

References 61 publications

AI et al.: Machines Are About to Change Scientific Publishing Forever

AI et al.: Machines Are About to Change Scientific Publishing Forever

Natural Selection Favors AIs over Humans

Complex QA and language models hybrid architectures, Survey

Contact Info

Product

Resources

About