Tech and Society Mod 9

Judy Kay- AI and education grand challenges

AI and Education (AIED) research - triggered by educational problems main goal - creating a system that functions the same as one-on-one teaching Interconnection - current AI technologies & information in use inform educational AI systems AIED challenges can be put back into regular AI research to help create new systems the learner model - has 4 elements model of learner, domain expertise, teaching expertise, and interfaces all aspects must must work in order to form an effective learning system Interconnection among 4 elements allows AIED to create interfaces to make learning experience very natural. E.g. tutor avatars programmed to identify with learner (adapt language, race, facial expressions and even gender) AI - utilized in education to create better learning experience for individual learners

W.D. Heaven - AI can simulate and economy... create fairer tax policy

AI and economic policy making getting the right taxation balance remains a challenge balancing productivity and income equality Salesforce - US business technology company developed AI Economist - a system that uses reinforcement learning seeks "to identify optimal tax policies for a simulated economy" AI Economist's simulated economy - mix of regressive & progressive tax policy system appears nimbler and more flexible In early stages AI Economist (or something similar) great potential to propose a better tax plan than humans

Zachary Davis- AI on the battlefield

Can assist both the tactical and strategic levels of war Tactical process large amounts of surveillance, intelligence, and reconnaissance data analyze a battlefield and decide quickly what should be done AI-controlled vehicles - explore terrain and different battlefield conditions exploring nuclear reliability of nuclear stockpile Strategic find specific targets given to them and attack that given target comb internet and hack into sites and domains in order to retrieve needed information Concerns enemies also have AI data can be flawed - AI can thus jump the gun in its decisions

Artificial Intelligence

Encompasses - machine learning, deep learning, neural networks, etc. technology with human-like cognitive activities learning, understanding, reasoning, and interacting AI Ecosystem - dominated by China, U.S.A and China (over 70 %) Broad And Growing Spheres Of Impact personal devices - computing human-computer interactions Businesses and organizations agriculture, e-government, banking, telecommunications, transportation, life and medical sciences, etc. Governments - smart cities, e-government, etc

Beth Singler- The AI creation meme

European Commission - strategic plan for future of (AI) generated a meme (AI Creation Meme) AI Creation Meme inspired by Michelangelo's "The Creation of Adam" artwork -nearly touching fingers of Adam and God highlights continuing mythologizing of relationship between human and the non-human other seen as a "'popular depiction of the supernatural'" the relevance of relative positions Michelangelo's - human and God European Commission's AI Creation Meme - Human and AI AI taking Adam's place in some renditions continuation of the theme of creation - humanity as creator giving life to the AI religious imagery "fits within a current post-secular context"

Noriko Arai- Can a Robot Pass a University Entrance Exam?

Impact of AI - only replace menial jobs and create enough new jobs for those replaced? Todai Robot - performed in the top 20% of students on Univ. of Tokyo entrance exam Todai Robot - cannot read or understand, but is great at searching & optimizing it excelled in factoid questions wrote better 600-word essay than most students used symbolic computation to pass complex math exam (beyond arithmetic) - in top 1% Did Todai Robot gain admission to University of Tokyo? No it does not understand meaning - disguises as if does top 20% performance could have gained admission to 60% of all other Japanese universities Is it problematic for Todai Robot to outperform most students? Need to thinking carefully and quickly on how humans can coexist with robots

AI and healthcare

Myriads of areas of impact to support work of care givers

Fei Jiang et al. AI in healthcare

Role and Areas assistive of human medical professionals handling healthcare data (patients' records, research findings, reported best practices, etc.) can facilitate better clinical decisions Two primary AI devices machine learning (ML) techniques for structured data natural language processing (NLP) - extract information from clinical notes, medical journals, etc. Potential early detection and diagnosis, treatment (reduction of errors) outcome prediction and prognosis evaluation Major disease areas - cancer, neurology and cardiology. Challenges remain for real-life deployment

AI and Military

changing the battlefield reducing human causalities (mostly of soldiers of deploying militaries) minimizing seriousness of war and its effects (?)

AI and Religion

concerns - morality and ethics of creating human-like machines intersection with apocalyptic forecasts Catholics - can plug into the Confession Chatbot (remove embarrassment in confessing?) Japan (2017) - Peppa (robot Buddhist priest) to conduct rituals can perform a funeral ceremony for $462 (human priest charges ___________)

Artificial Intelligence and the Economy

major sectors impacted - marketing & sales, supply chain management, logistics & manufacturing, transportation, automotive etc. competition with professional labor creation of labor markets shock - replacement of less skilled labor enhanced productivity - collaboration between workforce and A.I. programs new virtual workforce ensure improved quality control maximization of efficiencies - automation of routine tasks Enhance high standard of living at an optimal price - reskilling, upskilling, & adapting

AI and Education

transcending "pedagogical dark ages"? flipped lessons greater classroom interaction students within a class may progress at different speeds role of bots to expand facilitate recruitment and onboarding of students remind and cajole students to submit coursework; assess some of students' work, etc. Jill Watson - teaching assistant chatbot New generations of chatbots - set to be revolutionary great knowledge of the curriculum understand better approach to learning that suits each student best understand better which times of day we are most receptive, etc. acceleration & reduction of cost of learning - universal access to global classrooms