Maria Victoria Carro, Denise Alejandra Mester, Facundo Nieto, Oscar Agustin Stanchi, Guido Ernesto Bergman, Giovanni Franco Gabriel Marrafini, Trinidad Borrell, Guido Freire · 2026 · arXiv
DOI: 10.48550/arXiv.2610.28539 · open-access PDF
Como podemos efetivamente supervisionar sistemas de IA que superam a inteligência humana? A AI Safety via Debate propõe que dois sistemas de IA argumentem posições opostas e um juiz, seja um humano ou um modelo mais fraco, decida qual é o correto. Ao facilitar os supervisores em fornecer sinais de treinamento de alta qualidade, o debate visa incentivar o comportamento verdadeiro e seguro de sistemas de IA cujas capacidades excedem as nossas. No entanto, o julgamento humano não é neutro: as pessoas muitas vezes confiam em preconceitos que os sistemas avançados de IA podem aprender a explorar.
Machine translation (Meta M2M100); the English original follows.
How can we effectively oversee AI systems that surpass human intelligence? AI Safety via Debate proposes that two AI systems argue opposing positions and a judge, either a human or a weaker model, decides which is correct. By facilitating supervisors in providing high-quality training signals, debate aims to encourage truthful and safe behavior from AI systems whose capabilities exceed our own. However, human judgment is not neutral: people often rely on biases which advanced AI systems may learn to exploit. Drawing on insights from collective intelligence and social cognition, this paper inve
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.28539, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange