The Brightlands Smart Services Campus will be host to a full day of state of the art AI technology focused on applying ArtificiaI Intelligence in settings of strategic InterAction. Leading scientists and business experts will share AI developments. Also, latest AI experiments will be presented and project teams from the Techruption program will demonstrate the first prototypes developed in co-creation projects.
The program set-up chosen by the Business Intelligence and Smart Services Institute (BISS) triggers the dialogue between the business and academic AI community. Highlight of the day will be a keynote presentation by Prof. Tuomas Sandholm, Carnegie Mellon University, Pittsburgh. He has recently gained world-fame for his AI program Libratus, which is the first computer program that could win a Texas Hold’em poker tournament against world-class human players. Our second scientific keynote speaker is Prof. Stefan Kramer, Johannes Gutenberg University Mainz. He has gained high recognitions for his contributions to data mining, expressed by reward winning papers at several major data mining conferences. The business keynote will be given by Hans Martens and Manuel Winkler from Another Monday, a world-wide leading company in Robotic Process Automation.
09.30 - 10.00 Walk-in and registration
10.00 - 10.15 Welcome and opening notes
10.15 - 11.00 Faster (Algorithms), Higher Accuracy), Stronger (Computing Power): AI Algorithms and applications from Stream Mining to Deep Learning
Prof. Dr. Stefan Kramer, Professor of Data Mining, Institute of Computer Science, Johannes Gutenberg University Mainz
The current age of big data, data science and machine learning has brought new levels performance in almost all application areas. In the talk, I will give an overview of some important families of models and algorithms and discuss the benefits and limitations that still exist. The first part of the talk is devoted to stream mining and in particular the task of density estimation on data streams, which is one the most complex tasks that has been addressed in the data stream setting so far. The second part discusses deep learning approaches and applications to natural language processing (sentence modeling) and medicine (PET brain scans). I will discuss the issue of validating deep neural networks and possible ways of integrating deep nets with more knowledge-based approaches based on logics.
11.00 - 11.45 AI and the future of Robotic Process Automation (RPA)
Hans Martens, CTO, Another Monday
Manuel Winkler, Head of Technology and Project Management, Another Monday
12.00 - 13.00 Lunch break and networking
13.00 - 15.00 Parallel sessions (2 rounds, same sessions in each round, 1 round takes 50 min.)
Techruption AI demonstration
Stefano Bromuri and Rui Jorge Almeida, Assistant Professor, BISS Institute
Room: Variety 1.02
APG AI experiments
René Rateischak, Innovation Scout - AI Lead, APG Groeifabriek
Room: Velocity 1.03
Dare to disrupt!
Disruption starts with opening up for changes within your organization. These changes affect culture, processes, customer experiences and also, without mercy, IT. Wherever disruption and Digital Transformation show necessity, I will gladly participate. To start with, we will walk the path called Disruption together.
Arjen van Berkum, Chief Disruption Officer, Conclusion
Room: Access Point 1.59
From digital twins to Avatars-New Inhabitants in Smart City environments
Look around! We live in a world where more and more things are connected to internet This allows them to communicate and share information with each other, and… with us. In this presentation we would like to share our vision of tomorrow’s smart factories and smart cities, using the concept of Digital Twins. Regarding todays results in Machine Learning, and tomorrow’s promises on AI developments, we would like to talk about a new form of citizens: avatars. We will explain how we can combine a few contemporary technologies in order to let these avatars be your daily colleagues.
Wico Mulder, Lead Consultant Data Management, KPN
Room: Volume 1.01
Robotics and Service Design
Recent developments in robotic and ambient assistive technologies play a promising role in revolutionizing care provision to address the challenges of ageing populations and shortages of healthcare staff. Increasing Social Service Robots are entering healthcare systems and serving as central data and communication hubs. Yet issues such as privacy issues and threat to existing actors prevent their adoption. The network actors around the patient have their own views on benefits and risks of the robotic service, and can even be show-stoppers for the adoption decision. For example, when examining the introduction of social service robots in an elderly care setting, the final decision is influenced not only by the elderly (the focal actor), but also by family members, friends, GPs, nurses, and other professional service providers. This research introduces a design-based approach of value mapping of networks around elderly people to understand the varying perceptions within a network of users which hinder innovation acceptance. This network-conscious approach supports the introduction of new smart services with often disruptive effects, which are at the same time seen as both opportunities and threats. In this way, we offer advice to future design of meaningful robotic technology and its introduction in existing service networks.
Dominik Mahr, Associate Professor of Strategic Marketing & Innovation Management, Service Science Factory, Maastricht University
Martina Čaić, PhD Candidate, Service Science Factory, Maastricht University
Room: Server 1.27 + Switch 1.28
AI and Ethics (2 hours)
13.00 - 14.00 Taking the soft impacts of AI seriously
Humans have been delegating tasks to techniques (artefacts) since the dawn of humankind. But with Artificial Intelligence we not simply delegate physical or simple mental tasks to the machine, but rather our ability to evaluate and judge. This raises a new type of concerns. Indeed, AI may cause clear forms of harm, e.g. when an automated car crashes because of imperfections in the software. But the main concerns raised by AI are ‘softer’. For example, we worry that this technology may fundamentally change the ways we do things, the ways we meaningfully relate to one another, and even the ways we interpret ourselves. For example, if we delegate important legal or medical decisions to AI, who can be held accountable if things go wrong? How can we ensure that human biases are not built into AI systems? Or – maybe even more disconcerting – what if the ethical judgments of an AI system conflict with human ones? Does this mean that the system is biased, or that we are? If AI outperforms us in some mental tasks, how does this affect our self-understanding? This session will present some philosophical and ethical concerns raised by AI. As citizens, we will have to come up with answers to those questions, if AI is to improve our lives and our societies.
Tjsalling Swierstra, Head Philosophy Department, Maastricht University
14.00 - 15.00 Panel Discussion
- Manon van Beek, Country Managing Director, Accenture
- Laura Ahto, Chief Executive Officer, BNY Mellon European Bank
- Rudolf Müller, Scientific Director, BISS Institute
15.00 - 15.30 Coffee break and networking
15.30 - 16.30 The Algorithms Behind the Super-Human AI for Strategic Reasoning: Libratus Beats Top Professionals in No-Limit Poker
Prof. Dr. Tuomas Sandholm, Professor in Computer Science, Carnegie Mellon University Pittsburgh Pennsylvania
Poker has been a challenge problem in game theory, operations research, and artificial intelligence for decades. No program had been able to beat top players in large poker games. In January 2017, our AI, Libratus, beat a team of four top specialist pros in the main benchmark, heads-up no-limit Texas hold'em, which has 10^161 decision points. Libratus is powered by new algorithms in each of its three main modules: 1) computing approximate Nash equilibrium strategies before the event (i.e., computing a blueprint strategy for the entire game), 2) safe nested endgame solving during play (i.e., refining the blueprint strategy on the fly in parts of the game that are reached while preserving guarantees on exploitability), and 3) fixing its own strategy to play even closer to equilibrium based on what holes the opponents have been able to identify and exploit. The algorithms are domain-independent and have applicability to a variety of imperfect-information games such as finance, negotiation, strategic pricing, business strategy, strategic product portfolio optimization, auctions, political campaigns, cybersecurity, physical security, military, and steering evolution and biological adaptation (such as for medical treatment planning and synthetic biology). This is joint work with PhD student Noam Brown.
16.30 - 18.00 Drinks & bites “This could be your lucky number”
Your host of the day will be Mr. Red Sneakers, Gerrit Heijkoop
Tuomas Sandholm invites you to the AI | IA Conference at the Brightlands Smart Services Campus! Check out the video: