Dr. Bonnie Johnson
Naval Postgraduate School
Yonah Welker is a technologist and public expert for algorithms and policies, former advisor of the Ministry and authorities of AI and Data, visiting lecturer. Yonah Welker's contributions and work have been featured and added to the acts, reports and frameworks, featured by the White House PCAST, the World Economic Forum, OECD, UNESCO, the WHO, supported AI and Digital Acts, intergovernmental treaties, ontologies of assistive systems, AI, robotics, health, education, accessibility systems, programs and lectures.
Yonah Welker has made over a hundred appearances and published commentary to bring awareness to human-centered technologies and policies, provided tech and policy commentary for the public and authorities (e.g. AI and Data, Telecommunication, Economic and Social Development, technology and research institutions, consulting companies), served as an evaluator and expert to EU-Commission funded projects, contributed to research, development and adoption frameworks, MOOCs, including digital ecosystems, workplaces, educational and public spaces, curated and boarded Summits of AI for humanity, cross-national initiatives. Prior to it, Yonah Welker was co-founder of Hardwaretech think tank, cocreated tech ventures and projects, served as an innovator in residence, evaluator and expert for technology transfer and innovation ecosystems.
Metacognition for AI Safety
Artificial intelligence (AI) technologies are evolving rapidly and are poised to have far-reaching impacts in nearly every aspect of society around the globe. AI is an integral enabler of robotic systems—providing the means for sense-making, guidance, actions, and autonomy. AI-enabled decision aids offer insights into many of our most challenging problems including those that exhibit highly complex dynamics with unpredictable outcomes. AI can process vast amounts of data and information for detecting anomalies, pattern-matching, image recognition, and handling uncertainty. Along with AI’s exciting prospects of new and improved capabilities, comes inherent safety and usage risks involving unintentional behaviors, along with challenges in transparency, trust, bias, and ethics. AI’s cognitive capabilities enable them to reason about the world. The emerging discipline of metacognition is developing abilities for AI to reason about its own reasoning. Metacognition, in essence, offers guardrails to address challenges inherent to AI systems. And the beauty of metacognition is that it takes advantage of the very essence of AI systems - the “intelligence” – to address the challenges through self-knowledge. This presentation will introduce metacognition concepts, architectures, and current approaches. The presentation will describe current research efforts and approaches with an emphasis on metacognitive applications within the military domain.