Digital Governance & Artificial Intelligence
1 like1,269 views
Yannis Charalabidis gave a presentation on AI in governance. He discussed how AI is being used in various areas of public administration like service provision, back office processes, and policy design. He believes AI will have an enormous learning potential and impact areas like developing digital twins of cities, simulations to help with policy design and democracy, and developing truly smart city applications and agents. However, he notes that universal algorithms for complex societal simulations do not yet exist and more basic research is needed in areas like developing generic public sector agents and understanding systems.
Digital Governance & Artificial Intelligence
- 1. AI in Governance: Why, What and How Yannis Charalabidis Professor, Digital Governance University of the Aegean yannisx@aegean.gr
- 2. Your invited speaker, in one slide • Born in Athens, origin from the island Samos, Greece • Electrical & Computer Engineer (BSc/Msc), PhD in complex information systems, National Technical University of Athens • Trained in Entrepreneurship Development at UC Berkley and U Amsterdam • A certified professional coach, specializing in mentoring students and startupers Since 10 years old, I am having fun while working: • 10 years a part-time employee in the family business (B2B liquor store) • 10 years a researcher in R&D projects for businesses and governments • 10 years a manager in the software industry, at SingularLogic SA (Greece, The Netherlands, Poland, US, India and the world). • 10 years a professor at University of the Aegean, teaching and researching on digital governance and innovative entrepreneurship (while wandering around the world) Getting senior, steadily: • 2016: The 8th most prolific scholar on digital governance, worldwide • 2018: Among the 100 most influential people on digital governance, worldwide
- 3. “Rules of the game” • My aim : to help you find an answer and develop ten new questions • Have a question ? 1. You can put it in the chat, I will see while I speak 2. If relevant and burning, you take the floor and ask in person • Relax, focus, and enjoy !
- 4. Part A Introduction Digital Governance Fundamentals GOV1.0 Connected Governance GOV2.0 Open Governance GOV3.0 Smart / Disruptive Governance
- 5. Definition - #1 Digital (or Electronic) Governance refers to the utilization of Information and Communication Technologies (ICT’s), by Public Administration at National or Local Level, aiming at providing high quality digital public services to citizens and businesses (Wikipedia)
- 6. The 5 automation levels of digital services Level Level 0: No information on the service online Level 1: I can only find information online Level 2 : Can find information and forms online. Level 3: I can only apply for the service online Level 4: Fully digital, one stop, one second Level 5: Service starts by its own, is pre- completed for citizens Paper 100% 90% 70% 30% 0% 0% Time Days Days Hours Hours Minutes Seconds
- 7. More people are connected (Internet, Web, Social Media, IoT) More machine intelligence is available (AI, HPC) More data is available (Big data, Open data, Linked data, Blockchain) 2030 2010 Presented in Samos Summit, 2010
- 8. From Industry 1.0 to Industry 4.0
- 9. Government 1.0 Division of labor, rules of Economy, French Revolution, US Constitution Post Kingdom - democracies 1789 1870 1950 Government 2.0 Processes, Bureaucracy, Management Science, Workflow, Project Management, Logistics, Typewriter Post-Empire Democracies Government 3.0 Computers, Information Systems, Software, Service science, Interoperability Standards, Mobile devices Postwar democracies Government 4.0 Cyber-physical Systems AI, IoT, Smart solutions, Bid Data, Analytics, Policy modelling and simulation, stance classification, opinion mining, sentiment analysis New forms of Democracy ? dGOV1.0 dGOV2.0 dGOV3.0 Connected Open Smart / Disruptive Governance Governance Governance From Gov 1.0 to Gov 4.0
- 10. Definition - #2 Digital Governance is a social phenomenon, in which (a) the provision of services by the state to citizens and businesses, (b) the collaboration between administrations and citizens, (c) the decision-making and policy development, and the overall operation of the Public Sector, and mainly performed by digital means (Definitions / Gov30 project / www.gov30.eu)
- 11. The generations of Digital Governance dGOV 1.0 1990 + dGOV 2.0 2010 + dGOV 3.0 / 4.0 2020 + Main Goal Better Services Openness & Collaboration Societal problem solving, quality of living Main Method Connected Governance Open and Collaborative Governance Smart / Disruptive Governance Usual Application Level National Local Local to International Key Tool Portal Social Media Smart Phone Key Obstacle / Risk Public Sector Mentality Public Sector Mentality Problem Complexity Key ICT area Interoperability Open Data, e- participation AI, IoT, HPC Most needed disciplines, beyond ICT Management (BPR) Sociology, Marketing History, Psychology, Law, Political Science
- 12. The Domains of Digital Governance Digital Governance Digital Governance 1.0 (Service Provision) Digital Governance 2.0 (Open & Collaborative Governance) Digital Governance 3.0 / 4.0 (Smart & Disruptive) Service Automation and BPR Legal, Organisational, Semantic & Technical Interoperability e-Authentication Digital Signatures Service Portals Mobile applications Cloud Infrastructures e-Participation e-Voting Open Governmental Data (OGD) Linked Data e-Collaboration (Collaborative Service Design) Crowdsourcing Community Awareness Platforms Big Data-Driven Decision Making Artificial Intelligence in Governance Blockchain in Governance Internet of Things Policy modelling Societal Simulation Opinion Mining / Sentiment Analysis
- 13. Part B: AI in the public sector - The Status Time in Samos should now be: 17:10
- 14. Cases of AI in Governance, in European Union Member States (EC/JRC AI Watch Report, 2020) 1 3 2
- 15. 1 3 2 (EC/JRC AI Watch Report, 2020)
- 17. Estimated savings in US Federal Government (~2.5m employees) (Deloitte analysis, 2022) That’s 25% of the total workforce hours, of the 2,5M Federal Gov. employees Akshardham Temple, New Delhi : 30M hours to build Building 3 to 30 big temples, every year.
- 18. Part C: My View of the Future & Recommendations Time in Samos: 17:20 Greece is a Finalist ! @European eGovernment Awards, Malmoe 2009
- 19. My view of AI Autonomous Vehicles & Robots Soft-Bots & Chatbots Digital Twins Societal Modelling Self-adapting Simulation Automated Reasoners Intelligent Systems Large Corpus Understanding Decision Support
- 20. AI System Domain Key Tech Learning Potential Applicability My Opinion Chatbots Service Provision NLP / Machine Learning Supervi sed Medium Enough Let it happen Reasoning SoftBots / RPA agents Service Provision & BackOffice Process Automation NLP / Rule execution / WF’s Supervi sed Big Enough Go for it, try a few Other full learning SoftBots SP & BO & Policy Design NLP / (Deep) Machine learning / HPC Autono mous Enormous Partial / sectoral Invest in that City / Urban Digital Twins Policy Design, Smart City, Democracy IoT / Model based learning / HPC Autono mous, Deep Big Medium Big trend, big efforts Societal Simulation Systems Policy design, better democracy Model based & fuzzy reasoning / Agents Autono mous, Deep Big Sectoral, still no universal algorithms Will be a common “base”, you need it. My bets … 5 / 10
- 21. AI System Domain / Aim Key Tech Learning Potential Applicability My Opinion Vision Recognition Systems Security, Service Provision Image and Video Processing / Machine Learning Semi – autono mous Medium Enough Regulation testbed Robots and autonomous vehicles Service Provision, Internal Processes, Security, Kinetic Systems / Machine Learning Autono mous Big Enough Industry will lead Legislation Understanding Systems Better legislation and justice NLP / Machine Learning / HPC Legal Modelling Supervi sed Big Need some years of research Invest in that, if within your scope Really Smart City Apps & Agents Mobility, Health, Security, Work, Quality of Life IoT / Machine Learning / NLP Autono mous Enormous Enough That’s a shiny road The bots university Basic research, education, nextGen Bots NLP, Machine Learning Supervi sed Enormous Need some years My far guess Generic Public Sector Agents Basic Research, Machine Learning, NLP, Autono mous Enormous Need some years My far guess My bets … 10/10 + 1
- 22. Why ? – Example 1 AI in Legislation - Draft Legislation Checking (e.g. Digital-ready Legislation, AI- conforming legislation) - Citizen opinion processing, for collaborative decision making - Standardised Law production - Fully automated Law analysis, interrelation, translation, and publishing - Press/media information processing, for impact assessment - Citizens understand and navigate the law, can say their opinion during law design, can generate their own legislation checks (citizen science) - Parliaments easily consult other legislations, automate legislation checks, produce well codified laws. - MP’s control the key parts of the system, in parliament. - Law professionals have analytical tools to navigate 200 years of legal documents in 200 languages, for any issue - Governments and local administrations have access to codified law, easing digital transformation to rule based reasoning systems What is the goal How is done www.manylaws.eu
- 23. Why ? – Example 2 City Twin & Citizen Science - City digital twin for key infrastructures and layers - Citizen sensors for air and water quality, traffic, waste, etc - A constantly running simulator, on top of the digital twin (showing the next stages of several city systems, adapting as time goes) - Specialised projections and visualisations for every important city decision. Scenaria with possibilities. - A new city operation and design, with better decision making and fast response - Citizens become part of the research “infrastructure” of the city, acting as probes and running small labs, while being aware and changing attitudes - City council decisions are based on evidence, experimental results and simulation - Local administrations have a system with memory, operational capacity and forecasting abilities What is the goal How is done www.wecompair.eu
- 24. Want to contribute ? Want to have some fun, too ?