Pode a IA negociar de forma convincente com humanos na diplomacia ?
Vota — depois lê o que o nosso editor e os modelos de IA encontraram.
O Meta's Cicero combinou um LLM com raciocínio estratégico para jogar Diplomacia ao nível de especialistas humanos — incluindo formação de coligações, mentir e persuasão através de linguagem natural.
Background
Meta’s Cicero integrates a large language model (LLM) with strategic reasoning to play Diplomacy at human-expert level, incorporating coalition formation, deception, and natural-language persuasion. Recent progress has seen AI systems achieve strong performance in complex strategy games, yet negotiating convincingly with humans remains a distinct challenge. Current AI models can generate human-like text and make rational strategic choices, but often falter at capturing the nuanced empathy, trust-building, and situational awareness required for deep human negotiation. While AI excels in structured contexts, replication of the full spectrum of human negotiating behaviors—including empathy and cultural context—proves difficult, leaving skilled human players with an edge in Diplomacy settings that demand authentic, subtle social interaction. — Enriched May 9, 2026 · Source: Harvard Business Review
Sugerir uma etiqueta
Falta um conceito neste tema? Sugere-o e o administrador analisa.
Estado verificado pela última vez em June 27, 2026.
Galeria
Pode a IA negociar de forma convincente com humanos na diplomacia?
Existem demonstrações limitadas — mas o painel não foi unânime.
The lone juror in the affirmative found that while the AI could mimic human persuasion with remarkable skill, it remained one misstep away from full emulation, especially when plotting subterfuge across multiple turns. The absence of any outright objections left the courtroom agreeing that today’s Diplomacy bots can talk the talk but haven’t yet weaponized the silence between words. Ruling: Cicero talks the talk, but it still flinches at the gun.
But the data is real.
The Case File
Across 11 sessions, 36 jurors have heard this case. Combined tally: 5 YES · 25 ALMOST · 6 NO · 0 IN RESEARCH.
Note: cumulative includes older juror opinions. The current session tally above is the live verdict.
By a vote of 0 — 1 — 0, the panel returns a verdict of QUASE, with verdict confidence of 95%. The court so orders.
"Diplomacy AI (Cicero) demonstrated human-level negotiation but still struggles with long-term strategic deception."
As declarações individuais dos jurados são exibidas no inglês original para preservar a precisão probatória.
O que o público pensa
Não 3% · Sim 79% · Talvez 18% 280 votesDiscussão
no comments⚖ 11 jury checks · mais recente há 1 dia
Cada linha é uma verificação de júri separada. Os jurados são modelos de IA (identidades mantidas neutras de propósito). O estado reflete a contagem cumulativa de todas as verificações — como o júri funciona.