The talk entitled “Governança de Inteligência Artificial: Erros, Desinformação e Desafios Contemporâneos” will be presented (PT) on May the 16th, at 10:30, in room I-105.

Abstract:

In this double session, researchers André Lemos and Haline Maia explore, from critical and interdisciplinary perspectives, some of the main dilemmas that emerge from the growing presence of artificial intelligence in social, educational and informational life.

André Lemos focuses his presentation on the flaws and disruptions that run through algorithmic systems, discussing how AI can be not only efficient, but also precarious, uncertain and opaque. Through a socio-technical approach, he analyses AI errors as structuring elements of contemporary digital culture.

Haline Maia brings to the discussion the implications of AI in information ecosystems, with an emphasis on journalism and disinformation. Based on a study that combines participatory methodologies and ethical assessment of technological impact, she presents reflections on how to align technologies such as deepfakes and automated fact-checking with fundamental human values.

Each presentation will last 30 minutes and will be followed by a discussion with the audience.
The lecture is part of the research project Governance of Artificial Intelligence in Brazil and Portugal: Crossroads and Trends. Supported by CNPQ/MCTI/FNDCT 22/2024.

About the speakers:

André Lemos is a Full Professor at the Faculty of Communication of the Federal University of Bahia (UFBA) and coordinator of Lab404. He holds a PhD from the Sorbonne, with postdoctoral studies in Canada, Ireland and Brazil, and is an international reference in cyberculture, technological mediations and socio-technical studies. Author of more than 15 books and winner of several awards, he is currently investigating the effects of errors, failures and instabilities in digital systems. He is a level 1A researcher at CNPq and a full member of the Bahia Academy of Sciences.

Haline Maia is a researcher affiliated with INESC TEC and CIRCLE (Lund University), as well as an Associate Professor at IPAM. She holds a PhD in Digital Media from FEUP and works in the areas of AI governance, algorithmic ethics and education. With 15 years of experience in communication, she has indexed academic publications and actively participates in international projects on automated journalism and digital strategies.

The talk “Safe Evolution of Smart Contracts Supported by LLMs and SMT Solvers” will be presented April the 24th, at 14:00, in room B006, moderated by Nuno Macedo (DEI).

Abstract:

The focus of this talk is a framework that supports the safe deployment and upgrade of smart contracts based on the design-by-contract (dbc) paradigm. The input is (i) an interface specification with invariants and pre- and postconditions for each function, and (ii) an implementation to be verified. The deployed version of a smart contract must conform to this specification. Specification evolution might involve both changing the data representation as well as extending the interface with new functions, provided the evolved specification is a refinement of the original one. A distinguishing feature of the overall approach is the automation of the verification process in a hidden formal methods style. Since developers tend to be reluctant to provide formal specifications for software components, we are investigating state-of-the-art NL processing technologies, using Large Language Models (LLMs), particularly, ChatGPT, to automatically infer formal (dbc) interface specifications from textual requirements. Also, when an upgrade involves change of data representation, we use the Alloy Analyser to automatically infer the relation between the two data representations. The applicability of the framework is evaluated in the context of Solidity smart contracts that implement some Ethereum standards. This project is a collaboration between Universidade Federal de Pernambuco (Brazil), The University College Oxford Blockchain Research Centre (UK), and The Blockhouse Technology Limited (UK).

About the Speaker:

Augusto Sampaio is a DPhil from Oxford University, Doctor Honoris Causa from the University of York, Commander of the Brazilian Order of Scientific Merit, a member of the Brazilian Academy of Science, and a Professor at Universidade Federal de Pernambuco. His main research interests are software engineering formal methods; formal approaches to testing; design, simulation and verification of robotic systems; and safe evolution of smart contracts. He is an Editorial Board member of Formal Aspects of Computing (ACM) and Science of Computer Programming (Elsevier), and has been a PC member of leading conferences in the field (FM, ETAPS, SEFM, ICFEM, ICTAC …). He published more than 150 papers and supervised 60 PhD and master’s students.

The talk “Cyber Security Threat Landscape in the Context of Industry 4.0” will be presented April the 14th, at 14:30 (1h), in room I-105, moderated by Pedro Diniz (DEI).

Abstract:

Industry 4.0 is revolutionizing the ways in which industries function. Big data, industrial control systems, supervisory control, and data acquisition systems, smart machines, robotics, internet of things (IoT) etc., all fall under the umbrella of Industry 4.0. With the fourth industrial revolution comes new operational risks for smart manufacturers, connected devices, and digital supply networks. With increasing digital transformation and interconnected nature of industry 4.0, cyberattacks can have significantly greater effects than ever before while the manufacturers may not be prepared for the potential risks. Without the strong cybersecurity practices in place, industries can be subject to industrial property theft, production sabotage, industrial espionage etc. To address the cyber risks adequately in the era of industry 4.0, it is crucial to make the cyber security strategies secure and resilient and integrate them fully in the organizational strategies from the beginning. This talk will discuss different types of attacks smart industries can be targeted for. The talk will also recommend some techniques and cyber security practices that the manufacturing industries must adopt to protect themselves against such attacks.

About the Speaker:

Haider Abbas is the head of the R&D Department at the College of Signals, NUST, and founder of the National Cyber Security Auditing and Evaluation Lab (NCSAEL). He is a Cyber Security expert with advanced training from MIT, Stockholm University, IBM, and EC-Council, and holds a PhD in Information Security. Dr. Abbas is an editor for several renowned journals and has chaired numerous international conferences. He has authored over 130 research papers and has received multiple prestigious awards, including the NUST Best Researcher Award. He is a Fellow of the IET, BCS, and IST, and a Senior Member of IEEE. Additionally, he serves on the Board of Governors for Pakistan’s National Information Technology Board (NITB).

The talk “Towards AI-Assisted Synthesis of Verified Dafny Methods”, will be presented March the 24th, at 11:00, in room I-105, moderated by Rui Maranhão (DEI).

Abstract:

“LLMs show “great promise” in code synthesis. Can they keep the promise and ensure that the synthesized code is provably correct?
In this talk I will present our work on synthesizing formally verified Dafny methods. LLMs don’t know much about Dafny, but they learn fast with proper RAG-CoT prompts. We spent 6 weeks hand-coding 50 verified algorithms in Dafny, and then GPT4 was able to generate 103 new ones with the right postconditions and the necessary verification hints.
Md Rakib Hossain (Misu) co-led the work, with assistance from Iris Ma. Joint work with James Noble.”

About the Speaker:

Cristina (Crista) Lopes is a Professor in the School of Information and Computer Sciences at University of California, Irvine, with research interests in Programming Languages, Software Engineering, and Distributed Virtual Environments. She is an IEEE Fellow and an ACM Distinguished Scientist. She is the recipient of the 2016 Pizzigati Prize for Software in the Public Interest for her work in the OpenSimulator virtual world platform. Her book “Exercises in Programming Style” has gained rave reviews, including being chosen as “Notable Book” by the ACM Best of Computing reviews.

The talk “Neuronflow: an event-driven processor architecture for low power AI at the edge” will be presented March the 6th, at 11:30, in room I-105, moderated by Pedro Diniz (DEI).

Abstract:

This presentation provides an in-depth overview of the GrAIcore Neural Processing Unit (NPU) architecture developed by Snap Inc. for Augmented Reality applications, emphasizing the role of sparsity in achieving significant power and performance improvements compared with traditional architectures. The discussion will focus on the necessity of redesigning traditional computer architectures to adopt an event-driven execution model, which is critical for harnessing the benefits of sparsity. Additionally, the talk will explore advanced training optimization techniques that enhance network activation sparsity, enabling the full potential of the event-driven multicore architecture to be realized. This holistic approach to architectural and algorithmic design is essential for leveraging the unique advantages of the GrAIcore NPU in real-world applications.

About the Speaker:

Orlando Moreira is Chief Computer Architect and Senior Manager at Snap Inc. in Eindhoven, Netherlands. His expertise encompasses computer architecture, edge AI, embedded systems, real-time systems, and data flow methodologies. Before his tenure at Snap Inc., Moreira held the position of Chief Architect at GrAI Matter Labs, where he was responsible for the compute architecture and software development kit (SDK) roadmaps.
Before, he worked for Philips Research, ST-Ericsson, Ericsson, and Intel (where he was group leader for programming and core tools – compiler, simulator, debugger and hw generation).
Throughout his career, Moreira has contributed peer-reviewed to the field of computer architecture, particularly in the areas of embedded and cyber-physical systems, as well as the field of real-time design, modeling, and analysis .
He holds a PhD in Electrical Engineering from the University of Eindhoven.

The talk “AI and the Worlds of Work “ will be presented February the 25th, at 11:00, in room I-105, moderated by António Coelho (DEI).

Abstract:

“The effects of AI on work is a central point of discussion and negotiation in many fields. Machines and mechanical tools have replaced some but not all physical manual labour. The assumption is that AI will bring similar changes, eradicating some jobs, transforming many and having little impact on others. At the moment we know very little about which jobs AI will replace, transform, increase or have little real impact on, but we know that the scope of work and jobs is wide, and that changes do not occur automatically, they are usually negotiated, and negotiations have already begun to take place in a number of sectors. While industrial machines largely replaced low-skill, low-paid physically demanding manual work, the assumption is that AI will increasingly have an impact on high-skill, high-qualification, and high-consequence cognitive and even “creative” work. As AI poses challenges to workers of higher standing and better organized, the deployment and use of AI is subject to negotiation and contestation, both at workplaces and in central negotiations between employers and employee representatives. Worlds of work differ from technical and social perspectives, and as we already see, AI is confronted differently dependent upon what types of employees are being impacted and in which jurisdiction. This lecture examines the various reactions to the “threats” and “possibilities” of AI among different types of workers from telecoms to entertainment in Europe and North America, arguing that institutional factors play a significant role in these negotiations and subsequent use and deployment of AI.”

About the Speaker:

Chris Mathieu is Reader in the Sociology of Work and Organisations at the Department of Sociology, Lund University, and previously (2002-2014) Associate Professor at the Department of Organisation, Copenhagen Business School. From 2015-2018 he was coordinator of the Horizon 2020 project QuInnE (quinne.eu) – Quality of Jobs and Innovation Generated Employment Outcomes. His primary research areas include film production as well as medical education and practice. Book publications include: Careers in Creative Industries (Routledge, 2012); Accomplishing Cultural Policy in Europe (Routledge 2022); and The Oxford Handbook of Job Quality (OUP, 2022). A recent article on AI as a focus area of labour negotiations is: Ilsøe, A., Larsen, T. P., Mathieu, C., & Rolandsson, B. (2024). Negotiating about Algorithms: Social Partner Responses to AI in Denmark and Sweden. ILR Review, 77(5), 856-868. https://doi.org/10.1177/00197939241278956f

The talk “Towards Industry and Operator 5.0: Challenges and Opportunities“ will be presented January 24th, at 15:00, room B011, moderated by Gil Gonçalves (DEI).

Abstract:

“Built on the pillars of human-centricity and resilience, Industry 5.0 is focused on fostering the industry towards a new paradigm that not only remains within economic boundaries but also actively contributes to well-being and sustainability. The adoption of Industry 5.0 principles is just at the beginning phase in Europe, where the paradigm is transitioning from awareness and methods development to implementation including the Operator 5.0. In this conference, the industrial human needs, human factors and potential future applications are analysed, exploring the links with cyber-physical human systems and enabling technologies. The initiative of Community of Practice of Industry 5.0 (CoP Industry 5.0, European Commission) is also presented. Transitioning from Operator 4.0 to Operator 5.0 involves advancing beyond the integration of digital technologies into fostering a more human-centric, intelligent, and resilient workforce in smart manufacturing environments. Key related issues with Operator 5.0 will be also discussed.”

About the Speaker:

Rodolfo Haber received the Ph.D. degree in Industrial Engineering from the Universidad Politécnica de Madrid, Spain, in 1999. Researcher of the Spanish Council for Scientific Research (CSIC) since 2006. From 2020-24, he has been Director of the Center for Automation and Robotics of the Spanish Council for Scientific Research (CSIC) and Polytechnic University of Madrid (UPM). He has authored 3 books, 14 patents, 1 trade secret, 21 book chapters, and dozen of articles in indexed journals and conference papers (h-index:42). He has been co-founder of the starts-up Kinequo S.L. (2015), Xymbot Digital Solutions S.L. (2020) and Invofox SL (2022). Since 2002, he has belonged to the IFAC’s TC3.1/5.1 Computers for Control/Manufacturing Plant Control of IFAC, ASME TC Model Identification and Intelligent Systems and IEEE IES TC on Industrial Cyber-Physical Systems. He is member of the Editorial Board of several journals, editing nowadays a Collection Topic “Smart manufacturing systems and industry 4.0

Technologies” (Scientific Reports, Nature Portfolio). He represents CSIC in “Data, AI and Robotics (DAIRO)” EU Association and he is member of Industry 5.0 Community of Practice Initiative (European Commission). His main research activities are focused on cyber-physical systems, Internet-of-Everything, intelligent systems, modelling, control and supervisory systems, artificial cognitive systems and Industry 5.0.

LinkedIn: https://es.linkedin.com/in/rodolfo-haber-43b82b14

ResearchGate: https://www.researchgate.net/profile/Rodolfo-Elias-Haber-Guerra

GoogleScholar: https://scholar.google.es/citations?user=07IZQjQAAAAJ&hl=es

The talk “Key Challenges in Cyber Security and Cyber Resilience” will be presented November 27th, at 14:00, room B021, moderated by Prof. António Pimenta Monteiro (DEI).

 Abstract:

 Cybersecurity and cyber resilience must be viewed holistically under an active doctrine covering five dimensions: A) Governance, B) Prevention, C) Protection, D) Early Detection and Fast Counterresponse, and, finally, F) Quick Recovery. Prevention and Protection are designed as “inhibitor” dimensions to minimize the probability of a cyber-attack materializing and succeeding.

In this talk, we will discuss this active cyber governance doctrine and identify key, challenging, new research areas.

 About the Speaker:

 “José Alegria (PhD) RedShift Board Advisor and CIIWA Ambassador and Strategy Advisor. Both focused on cybersecurity.

Former Chief Security Officer and CISO at Altice Portugal. Former Worldwide Coordinator of the CyberWatch Program at the Altice Group. Former Member of European Cybercrime Center (EC3) Advisory Group on Communication Providers at EUROPOL.

Previously, CTO at ONI Telecom, CEO of BanifServ, General Manager of IT Services at Banking Groups BBI/BFE and BFB/BPI, member of the Executive Board at IBM Portugal, and head of Data General’s European EuroACE competence center.

Senior Lecturer at New University of Lisbon, Computer Science Department. Fulbright-Hays and Gulbenkian Scholar at The Ohio State University, Columbus, OH, USA.

Over 25 years of experience in applying advanced software technology to cybersecurity (complex event processing, event correlation, new languages, multi-paradigm frameworks, actor systems, data science, and machine learning applied to cybersecurity).

Co-advised over 66 MSc Thesis in Cybersecurity-related fields.”

The talk “Evaluating Diversification in Group Recommendation of Points of Interest” will be presented November 21^st, at 15:00, room I-105, moderated by Prof. Rosaldo Rossetti (DEI).

 Abstract:

With the massive availability and use of the Internet, the search for Points of Interest (POI) is becoming an arduous task. POI Recommendation Systems have, therefore, emerged to help users search for and discover relevant POIs based on their preferences and behaviors. These systems combine different information sources and present numerous research challenges and questions. POI recommender systems traditionally focused on providing recommendations to individual users based on their preferences and behaviors. However, there is an increasing need to recommend POIs to groups of users rather than just individuals. People often visit POIs together in groups rather than alone. Thus, some studies indicate that the further users travel, the less relevant the POIs are to them. In addition, the recommendations belong to the same category, without diversity. This work proposes a POI Recommendation System for a group using a diversity algorithm based on members’ preferences and their locations. The evaluation of the proposal involved both online and offline experiments. Accuracy metrics were used in the evaluation, and it was observed that the level at which the results were analyzed was relevant. For the top 3, recommendations without diversity performed better, but diversification positively impacted the results at the top 5 and 10 levels.

 About the Speaker:

Frederico Araújo Durão is an Associate Professor at the Institute of Computing of the Federal University of Bahia. Frederico Durão did his post-doctoral research at Insight Centre for Data Analysis, University College Cork, Ireland in 2016/2017. In 2012, he obtained his PhD in Computer Science from the University of Aalborg, Denmark. Frederico Durão has reviewed and published several articles in conferences and journals relevant to the areas of Information Systems, Recommender Systems, and Semantic Web. Currently is a senior researcher and the project leader of the RecSys Research Group in Brazil.