基本的毒性是指在线多人游戏中玩家表现出的粗鲁或侮辱性行为,这是困扰游戏界永恒的问题。 《 Dota 2》和《 Counter Strike:全球攻势》是遭受此问题困扰的两个主要游戏,与其他任何在线游戏相比,它们通常被认为是最具毒性的在线社区。 FACEIT是Dota 2,CS GO,英雄联盟等游戏的最大竞赛对接平台之一,已与Alphabet合作’的Jigsaw Labs通过识别服务器中的有毒玩家并惩罚他们的不良行为来遏制在线游戏中的毒害。


拼图实验室是Alphabet Inc(Google’的母公司),并通过将技术解决方案应用于全球挑战来充当技术孵化器。面对它’与Google的合作伙伴关系披露了名为的Admin人工智能系统的开发‘Minerva’, the first of its kind to make its way to esports. This AI system makes use of machine learning to scan through thousands of chat messages and quickly identifies toxic texts. The AI then critically analyzes the breach and immediately notifies the abuser in a matter of seconds making him/her aware of their wrongdoings. 密涅瓦 can also effectively ban repeat offenders, the ones who continue to exhibit toxic behavior after being notified several times in a short span of time. The AI system does not need human moderation and can reportedly function on its own.


测试阶段将于8月在FACEIT上开始’s servers and so far, 密涅瓦 has analyzed 200,000,000 chat texts and marked over 7,000,000 as toxic. The system issued over 90,000 warnings and 20,000 bans for verbal abuse and spam messaging. FACEIT as an additional measure, enforces phone number verification for accounts marked with toxic behavior, smurfing and spamming. Since the application of the system, FACEIT saw a 20% decline in toxic behaviour owing to the threat of instant punishment. Hence we can say that it definitely works.