ミッション Mission

AIと共存する人間中心の社会の実現
Realize a society centered on people coexisting with AI

Our aim is to realize a society centered on people coexisting with AI. We will closely coordinate the development of talented individuals and the implementation within society of specialized research that makes practical use of AI.

    • 信頼性を向上させた次世代大規模言語モデルの構築および学内外への提供
      Construction of a next-generation large-scale language model with improved reliability and provision of the model both inside and outside the university
    • 研究成果に基づく人材育成コースの提供
      Provide human resource development courses based on research results
    • シンポジウムやワークショップの開催による先端的研究成果の共有
      Sharing advanced research results by holding symposiums and workshops
    • 各専門領域におけるAI適応例の蓄積、学内のAI適応コンソーシアムの形成による適応法および適応可能性の公開と普及
      Accumulation of examples of AI adaptation in each specialized area, and public disclosure and dissemination of adaptation methods and their adaptability through the formation of an AI adaptation consortium within the university
    • 国内外のAI研究拠点との連携とハブ機能の実現、産官学連携の推進
      Collaboration with domestic and international AI research centers and realization of hub functions, and Promotion of industry-government-academia collaboration
    • 学内を含む教育機関向けの文章生成AIを用いた教育活動の探索および普及
      Exploration and dissemination of educational activities using text generation AI for educational institutions, including on-campus
    • 生成系AIの積極的活用社会における利活用ルールの提案
      Propose rules for the active use of generative AI in society

 

PHILOSOPHY

AI共生社会実現のための課題を解決する
Resolving Issues to Achieve an AI-Symbiosis Society

■　AI倫理 / AI Ethics

  • 人間とAIが共存する社会は、人間の尊厳が尊重され、多様な背景を持つ人々がそれぞれの幸せを追求でき、持続性がなくてはならない
    A society in which humans and AI coexist must respect human dignity, allow people of diverse backgrounds to pursue their own happiness, and be sustainable
  • 人間中心／教育・リテラシー／プライバシー確保／セキュリティ確保／公正競争確保／公平性、説明責任および透明性／イノベーションの原則
    People-centered / Education and literacy / Ensuring privacy / Ensuring security / Ensuring fair competition / Fairness, accountability and transparency / Principles of innovation

■　AIセーフティ /AI Safety & Security

  • 「事故」に対する頑健性、安全性
    Robustness and safety against “accidents”
  • 害を及ぼす副作用を生まない／報酬ルール悪用の回避／監視の適切な配分／安全な外挿／学習環境からの偏移に対する頑健性
    Does not produce harmful side effects / Avoidance of reward rule abuse / Appropriate allocation of monitoring / Safe extrapolation / Robustness to deviations from the learning environment

■　AIアライメント / AI Alignment

  • AIの能力を人間の意図した目的に合致させる
    Match AI capabilities to intended human objectives
  • AIの振る舞いを人間の考える倫理原則に合致させる
    Aligning AI behavior with human-oriented ethical principles
TOP