KI hat Schattenseiten und birgt die Gefahr, Menschen in unserer eh schon vereinsamten Gesellschaft weiter zu isolieren. KI bringt aber die Chance, unserem Planeten etwas zurückzugeben. Unser Fortschritt und Technologielevel wären heute nicht da, wo sie sind, ohne dass wir so viel von unserer wunderbaren Erde genommen hätten. Jetzt ist ein genialer Zeitpunkt zurückzugeben, und das passiert auch schon.
In meinem Vortrag erzähle ich euch, wie KI zu den 17 UN-Zielen für nachhaltige Entwicklung beiträgt.

Zielpublikum: Keine Einschränkung: Architekt:innen, Entwickler:innen, Projektleiter:innen, Manager:innen, Entscheider:innen, ...
Voraussetzungen: Keine Vorkenntnisse erforderlich
Schwierigkeitsgrad: Anfänger

Extended Abstract:
KI hat Schattenseiten und birgt die Gefahr, Menschen in unserer eh schon vereinsamten Gesellschaft weiter zu isolieren. KI bringt aber die Chance, unserem Planeten etwas zurückzugeben. Unser Fortschritt und Technologielevel wären heute nicht da, wo sie sind, ohne dass wir so viel von unserer wunderbaren Erde genommen hätten. Jetzt ist ein genialer Zeitpunkt zurückzugeben, und das passiert auch schon.
In meinem Vortrag mache ich auf die Gefahren und Schattenseiten von KI aufmerksam, von der algorithmisch perfekten Diskriminierung über die Liebesbeziehung mit Robotern, die organisierte Kriminalität mit ausgeklügelter KI bis hin zum riskanten Spiel mit dem Fortbestand der Menschheit.
Weiter erzähle ich euch, wie KI zu den 17 UN-Zielen für nachhaltige Entwicklung beiträgt. Es geht um Chancen, die die rasante Entwicklung künstlicher Intelligenz für unsere Zukunft bringt: Wie leisten Ozeandaten und KI Artenschutz? Wie bekämpft KI die Flussblindheit in Afrika? Wie rettet KI den Amazonas, Bienenstöcke und vieles mehr?

Mehr Inhalte dieses Speakers? Schaut doch mal bei sigs.de vorbei: https://www.sigs.de/autor/eldar.sultanow

While AI systems differ in some points from "traditional" systems, testing them does not have to be more difficult - knowing the right questions to ask will go a long way. In this talk we will:

• Arm you with a checklist of questions to ask when preparing to test an AI system
• Show you that testers and data scientist have common ground when testing AI systems

Keep calm and test on - AI systems are not that different from "normal" systems.

Target Audience: Testers, Data Scientists, Developers, Product Owners, Architects
Prerequisites: Basic knowledge of software testing
Level: Advanced

Extended Abstract:
If you're a tester about to test your first AI system, or wanting to move into that area, you're probably wondering how you can prepare for the role. While we usually do not deal with complexity in the magnitude of Large Language models like chatGPT, AI systems still seemingly offer different challenges than "traditional" systems.
You're not the first person to deal with these questions. In fact, a group of us got together to explore it in more detail. Is there a general framework of questions that testers can use to help develop a quality strategy for systems that use AI? We wanted to see if we could design one. To this end, we got together a group with diverse roles: tester, test architect, data scientist, project lead and CEO.
Join us in this talk to hear how we approached the task and what our results are, including an example of using our checklist of questions to analyse a system that uses AI. Along the way we also addressed questions like "What is the role of a tester in such projects?" and "How much math do I need?" - we'll talk about those discussions, too. We encourage participants to use our checklist and give us feedback on it!

Security engineering from TARA and security requirements to security testing demand mechanisms to generate, verify, and connect the resulting work products. Traditional methods need lots of manual work and yet show inconsistencies and imbalanced tests. Generative AI allows novel methods with semi-automatic cyber security requirements engineering, traceability, and testing. In this industry presentation, we show two promising approaches with NLP and transformers and how to embed them into an industry-scale security pipeline from TARA to test.

Target Audience: Test Engineers, QA Experts, Security Experts, Requirements and Systems Engineers
Prerequisites: Some background on security and testing. We will hands-on introduce the AI methods.
Level: Advanced

Extended Abstract:
Security engineering from TARA and security requirements to security testing demand mechanisms to generate, verify, and connect the resulting work products. Traditional methods need lots of manual work such as for traceability and yet show little impact when looking at the many inconsistencies and imbalanced tests. NLP especially transformers allow novel methods with semi-automatic cyber security requirements engineering, traceability, and testing.
We focus here on using generative AI with NLP because they can support the methods described in the standard while there is no need to change the form of representation from what is required by cybersecurity standards and respective stakeholders. Especially the use of Large Language Models (LLM) for text generation, aggregation, and classification has recently proven promising to improve the efficiency and effectiveness of security analysis and tests.
Grey Box Penetration Testing is an approach where only publicly available information is used to perform an attack on the SUT. This often requires massive research effort. Threat catalogs were known and often used threats are recorded can increase the performance while testing. To provide additional aid we are currently working towards building an AI-supported threat catalogue. Therefore, we use a special transformer model which is specialized in searching and summarizing information. When fed with known information about the SUT this model searches all available databases like CVE or CAPEC, previously recorded attack patterns, and other contextual information available and gives the penetration test engineer an initial idea of how to approach an attack on the SUT.
Using the AI to generate both grey and white box attack paths is an approach to check how much information about the system or components such as libraries and dependencies which are used in the SUT are available. Having introduced these methods to the security life-cycle, we will in the next step better integrate the tools. This will facilitate a swift turn-around upon changes in an agile delivery pipeline and thus achieve consistency from TARA to security requirements and (regression) test cases.
Vector together with the University of Stuttgart has developed transformers and generative AI-based methodologies for the specification and validation of cybersecurity requirements with the goal to increase productivity and quality.
In this industry presentation, we practically show how generative AI can scale into an industry-scale security pipeline.

Mehr Inhalte dieses Speakers? Schaut doch mal bei sigs.de vorbei: https://www.sigs.de/autor/christof.ebert

The introduction of ChatGPT, CoPilot X brings in a lot of hype over developer experiences, especially documentation. But are we ready to chat with our documentation, instead of reading, using these tools? How can we, as maintainers, leverage these tools to offer a better experience in documentation for developers as our users? Join my talk and let's find out.

Target Audience: Engineers, Developers
Prerequisites: Programming
Level: Advanced

Discover the transformative power of GenAI in software testing. This lecture showcases a powerful GenAI-powered test framework that enhances testing efficiency. Learn how GenAI analyzes applications to generate automated test cases, uncover hidden defects with generative AI's random exploratory tests. Experience AI-powered peer reviewers for code analysis and quality evaluations. Explore Smart Report AI, providing comprehensive analysis and insights into test execution, results, and defects. Join us to revolutionize your software testing with GenAI.

Target Audience: Quality Engineers, Architects, Developers, Project Leader, Managers
Prerequisites: Basic knowledge on modern software development
Level: Advanced

Extended Abstract:
In the rapidly evolving landscape of software development Generative AI holds immense potential to revolutionize the field of software testing. This lecture aims to explore the transformative capabilities of GenAI in software testing by showcasing the powerful GenAI powered test framework and its key components.
The lecture begins by introducing the main framework, which combines the power of GenAI and automation to enhance the efficiency and effectiveness of software testing. Attendees will learn how GenAI can analyze the application under test and generate automated test cases based on a predefined test framework. These test cases can be executed locally or seamlessly integrated into DevOps processes, allowing for efficient and comprehensive testing of applications.
Furthermore, the lecture delves into the utilization of generative AI for the generation of random exploratory tests. By leveraging the capabilities of GenAI, testers can uncover hidden defects and vulnerabilities that may go unnoticed with traditional testing approaches. This demonstration highlights the innovative potential of GenAI in driving comprehensive test coverage.
Additionally, the lecture showcases the Smart Peer AI component of the framework. Attendees will discover how AI-powered peer reviewers can anticipate functional and non-functional issues, providing insightful code analysis and quality evaluations. By incorporating specific coding best practices, architectural standards, and non-standard attributes, Smart Peer AI enhances code quality and accelerates the development process.
Finally, the lecture concludes by presenting the Smart Report AI, an AI-powered solution that analyzes test reports generated from automated testing processes. This component provides a comprehensive analysis of test execution, environment configuration, test results, defect severity automated analysis, reproducibility steps, and suggestions for fixes. Smart Report AI empowers testers with valuable insights and facilitates effective decision-making in the testing and development lifecycle.

In today's economy, creating intelligent customer experiences is a key differentiator for organizations looking to compete and gain a competitive advantage. Use of AI and especially Generative AI became more prevalent in the business world. We will discuss some of the work we did on creating an AI-Driven CX Platform that offers data management, Customer360 views, personalization and chatbots infused with Generative AI, and advanced security features. We will also discuss practical use cases and outcomes of our approach.

Target Audience: Managers, Decision Makers, Leaders
Prerequisites: None
Level: Advanced

This session will introduce embedding vectors and their use in artificial intelligence. It will illustrate how these constructs can be effectively utilized in enterprise AI solutions, specifically in conjunction with prompt engineering. Rainer Stropek will present practical demonstrations using Microsoft's Azure Cloud and OpenAI's ChatGPT 4 model, showcasing real-world application scenarios and potential business benefits. Attendees will gain insights into emerging AI trends and practices in enterprise contexts.

Target Audience: Architects, Developers, Data Scientists
Prerequisites: Basic knowledge about LLMs, ability to read source code
Level: Advanced

Extended Abstract:
This session aims to demystify the technical concepts of embedding vectors and their integration into the Artificial Intelligence (AI) world, with a specific focus on enterprise applications.
We provide an accessible introduction to embedding vectors – the mathematical constructs that translate complex data types into a format that machine learning algorithms can comprehend. Attendees will grasp the foundational principles of embeddings and their pivotal role in enhancing the capabilities of AI algorithms.
Next, we delve into the realm of prompt engineering. Here, we explore how the strategic crafting of prompts can steer AI models, such as OpenAI's ChatGPT 4, toward generating more precise and contextually accurate responses. We'll explain the process and strategies involved in prompt engineering, offering a roadmap for businesses to adopt these practices effectively.
Rainer Stropek will demonstrate these principles in a practical setting using Microsoft's Azure Cloud. We will walk through real-life scenarios, showing how enterprises can implement such technologies.
The session targets AI enthusiasts, software architects, developers, and data scientists. By the end of this session, participants will have gained a better understanding of these AI concepts, and how they can be harnessed.

Mehr Inhalte dieses Speakers? Schaut doch mal bei sigs.de vorbei: https://www.sigs.de/autor/rainer.stropek

In the evolving AI landscape, the EU AI Act introduces new standards for assuring high-risk AI systems. This presentation will explore the tester's role in navigating these standards, drawing from the latest research and from our experiences with an Automatic Employment Decision System, a high-risk AI. We'll discuss emerging methodologies, conformity assessments, and post-deployment monitoring, offering insights and practical guidance for aligning AI systems with the Act's requirements.

Target Audience: QA Professionals, AI Engineers/Architects, Business Leaders, POs/PMs, Policy Makers
Prerequisites: Basic Understanding of AI, Familiar with Testing, Awareness of EU AI Act, Interest in AI Asurance
Level: Advanced

Extended Abstract:
As the landscape of Artificial Intelligence (AI) rapidly evolves, the upcoming EU AI Act is set to introduce a new paradigm for assuring high-risk AI systems. This session will allow participants to delve into the pivotal role of testers in this context. We will decode the complexities of the Act and get to see how fostering the Act will ensure robust, transparent, and ethically aligned AI systems.
Drawing on recent research and my own experience in testing high-risk AI systems, I will discuss the emerging methodologies for testing high-risk AI, including explainability methods, robustness testing, and fairness testing. Together, we will also explore the Act's emphasis on conformity assessments and post-deployment monitoring, highlighting the tester's role in these processes. Participants will gain a unique behind-the-scenes look at how we have gone from chaos to order in testing an Automatic Employment Decision System, a high-risk AI.
Joining this session will equip participants with valuable insights and practical guidance on aligning AI systems with the EU AI Act. This is a must-attend for testers, AI developers, and business leaders alike who are navigating this new frontier, exploring the challenges, opportunities, and future directions in the assurance of high-risk AI systems. By the end of the session, participants will be better prepared to face the challenges posed by the EU AI Act and will have a clear understanding of the future directions in the assurance of high-risk AI systems.

Mehr Inhalte dieses Speakers? Schaut doch mal bei sigs.de vorbei: https://www.sigs.de/autor/andrei.nutas

One of the fundamental challenges for machine learning (ML) teams is data quality, or more accurately the lack of data quality. Your ML solution is only as good as the data that you train it on, and therein lies the rub: Is your data of sufficient quality to train a trustworthy system? If not, can you improve your data so that it is? You need a collection of data quality “best practices”, but what is “best” depends on the context of the problem that you face. Which of the myriad of strategies are the best ones for you?

Target Audience: Developers, Data Engineers, Managers, Decision Makers
Prerequisites: None
Level: Advanced

Extended Abstract:
This presentation compares over a dozen traditional and agile data quality techniques on five factors: timeliness of action, level of automation, directness, timeliness of benefit, and difficulty to implement. The data quality techniques explored include: data cleansing, automated regression testing, data guidance, synthetic training data, database refactoring, data stewards, manual regression testing, data transformation, data masking, data labeling, and more. When you understand what data quality techniques are available to you, and understand the context in which they’re applicable, you will be able to identify the collection of data quality techniques that are best for you.

Expanding horizons has many facets. It means taking advantage of new opportunities that arise from technical progress, such as Large Language Models, or societal challenges like Sustainability. Expanding horizons also means taking responsibility. AI and data analytics have a direct impact on our future life, good and bad. Expanding horizons also means reflection on existing practice. We have perhaps forgotten the benefits of structured monoliths, or have sometimes overdone it with agility, which suggests a critical and learning retrospective.

Moderation: Frank Buschmann
Panelists: Isabel Bär, Zoe Lopez-Latorre, Carola Lilienthal, Xin Yao

Target Audience: Software Practitioners
Prerequisites: None
Level: Advanced

Extended Abstract:
The motto of OOP 2024 has many facets. Expanding horizons means understanding and taking advantage of new opportunities that arise from technological progress or societal challenges. For example, on Large Language Models, Sustainability, and the Metaverse. Expanding horizons also means taking responsibility and not blindly applying new technologies. For example, AI and data analytics have a direct impact on our future life and social interaction. With all the consequences, good and bad. But broadening horizons also means reflecting on existing technologies and practices. In the course of the euphoria about microservices, for example, we have perhaps forgotten the advantages of the structured monolith too much, or have sometimes overdone it with agility. A critical and learning retrospective seems appropriate.
In this panel, we will examine the various aspects of the motto of OOP 2024 to give us all meaningful guiding thoughts for the exciting journey to expanding our horizons.

During the talk, we'll dive into the historical context of Generative AI and examine their challenges. From legal compliance to fairness, transparency, security, and accountability, we'll discuss strategies for implementing Responsible AI principles.
It's important to note that the landscape for AI-driven products is still evolving, and there are no established best practices. The legislative framework surrounding these models remains uncertain, making it even more vital to engage in discussions that shape responsible AI practices.

Target Audience: Decision Makers, Developers, Managers, Everyone - AI-driven products require cross-functional teams
Prerequisites: None
Level: Basic

Extended Abstract:
Foundation models like GPT-4, BERT, or DALL-E 2 are remarkable in their versatility, trained on vast datasets using self-supervised learning. However, the adaptability of these models brings forth ethical, socio-technical, and legal questions that demand responsible development and deployment.
During the talk, we'll delve into the history of AI to better understand the evolution of generative models. We'll explore strategies for implementing Responsible AI principles, tackling issues such as legal compliance, fairness, transparency, security, accountability and their broader impact on society.
It's important to note that there are currently no established best practices for AI-driven products, and the legislative landscape surrounding them remains unclear. This underscores the significance of our discussion as we collectively navigate this emerging field.

Are Large Language Models (LLMs) sophisticated pattern matchers ('parrots') without understanding or potential prodigies that eventually surpass human intelligence? Drawing insights from both camps, we attempt to reconcile these perspectives, examines the current state of LLMs, their potential trajectories, and the profound impact these developments have on how we engineer software in the years to come.

Target Audience: Developers and Architects
Prerequisites: A basic understanding of Large Language Models is helpful but not required
Level: Basic

Extended Abstract:
Large Language Models (LLMs) are complex 'black box' systems. Their capabilities remain largely mysterious, only beginning to be understood through interaction and experimentation. While these models occasionally yield surprisingly accurate responses, they also exhibit shocking, elementary mistakes and limitations, creating more confusion than clarity.
When seeking expert insights, we find two diverging perspectives. On one side, we have thinkers like Noam Chomsky and AI experts such as Yann LeCun, who view LLMs as stochastic 'parrots' — sophisticated pattern matchers without true comprehension.
In contrast, AI pioneers like Geoffrey Hinton and Ilya Sutskever see LLMs as potential 'prodigies' — AI systems capable of eventually surpassing human intelligence and visionaries like Yuval Noah Harari that view LLMs as substantial societal threats.
Regardless of whether we see LLMs as 'parrots' or 'prodigies', they undeniable are catalyzing a paradigm shift in software engineering, broadening horizons, and pushing the boundaries of the field.
What are the theories underpinning these experts' views? Can their perspectives be reconciled, and what can we learn for the future of software engineering?
To answer these questions, we examine the current capabilities and developments of LLMs and explore their potential trajectories.

Artificial Intelligence (AI) has become integral to software development, automating complex tasks and shaping this field's future. However, it also comes with challenges. In this talk, we explore how AI impacts current software development and possibilities for the future. We'll delve into AI language models in programming, discussing pros, cons and challenges. This talk, tailored to both supporters and skeptics of AI in software development, doesn't shy away from discussing the ethical obligations tied to this technology.

Target Audience: Software Engineers, Architects and Project Leaders in an enterprise environment
Prerequisites: Basic Understanding of Software Development
Level: Advanced

Extended Abstract:
The integration of Artificial Intelligence (AI), especially AI language models, is fundamentally reshaping the landscape of software development. This transformative technology offers exciting possibilities, from enhancing developer productivity to simplifying complex tasks. However, it is crucial to acknowledge and address the inherent challenges that accompany these advancements.
In this session, we delve into AI's role in software development, emphasizing both the potential benefits and potential drawbacks of AI. We will also examine the ethical implications of this technology.
By the end of the session, attendees will have a comprehensive understanding of AI's role in software development, reinforced by practical demonstrations. Participants will gain insight and skills for their daily roles as well as strategies for adopting and applying AI within their software development context.
As AI continues to permeate software development, understanding its potential and preparing for its challenges has become vital. My session, directly tied to the conference theme of "Expanding Horizons", provides unique, practical insights to help attendees navigate an AI-integrated future. Join me in encouraging the responsible and effective use of this transformative technology.

Tauchen Sie in die faszinierende Welt von ChatGPT ein! Dabei werden die Sprecher auf die Grundlagen von GPT und auf die besonderen Fähigkeiten von einem der aktuell leistungsfähigsten textbasierten Dialogsysteme eingehen.
Ein wichtiger Aspekt ist darüber hinaus, das Verständnis für Prompt Engineering zu fördern. Durch die Optimierung von Prompts lässt sich die Qualität der Antworten von ChatGPT drastisch verbessern. Wir behandeln einige bewährte Verfahren für die Konstruktion von Prompts, wie z.B. das Few-Shot oder Chain-of-Thought Prompting.

Zielpublikum: Architekt:innen, Entwickler:innen, Projektleiter:innen, Manager:innen, Entscheider:innen, Benutzer:innen
Voraussetzungen: Keine
Schwierigkeitsgrad: Anfänger

Extended Abstract:
Sind Sie bereit, Ihre Fähigkeiten im Bereich Prompt Engineering auf ein neues Level zu bringen?
In dieser Präsentation tauchen Sie in die faszinierende Welt von ChatGPT ein. Dabei werden die Sprecher auf die Grundlagen von GPT und auf die besonderen Fähigkeiten von einem der aktuell leistungsfähigsten textbasierten Dialogsysteme eingehen.
Ein wichtiger Aspekt ist darüber hinaus, das Verständnis für Prompt Engineering zu fördern. Durch die Optimierung von Prompts lässt sich die Qualität der Antworten von ChatGPT drastisch verbessern. Wir behandeln einige bewährte Verfahren für die Konstruktion von Prompts, wie z.B. das Chain-of-Thought oder das deklarative Prompting.
Sie gewinnen in dieser Präsentation ein tieferes Verständnis von ChatGPT. Praktische Tipps und Live-Demonstrationen zur Verbesserung von Prompt Engineering runden den Vortrag ab.

Klimawandel, Energiewende, ChatGPT und der Krieg verändern die Art, wie wir die Zukunft sehen. Neue Paradigmen empfinden wir allerdings als Stress und igeln uns ein. O-Ton von Ingenieuren: "Wir haben Benzin im Blut." Geisteswissenschaftler: "ChatGPT führt in die Hölle der Roboterherrschaft!" Entwickler: "ChatGPT macht Fehler!" Kulturverweigerer vergeuden wertvolle Zeit mit "Fluch oder Segen?"-Diskussionen auf niedrigstem Niveau. Ambivalenzfeste Führung durch den Sturm ist gefragt! Wie wäre es mit der Beherrschung "The Art of Prompting!"?

Zielpublikum: Für alle!
Voraussetzungen: Interesse für Transformation im Wandel
Schwierigkeitsgrad: Fortgeschritten

Extended Abstract:
Gerade die IT sieht sich vor Herausforderungen. Sie muss damit umgehen, dass Chat-GPT durchschnittlich gut coden kann. Sind Sie bereit dafür? Verlieren Sie Ihren Arbeitsplatz oder sind Sie resilienter Starentwickler? Wir stehen in einem Wettlauf des Menschen mit der Maschine, nehmen Sie es hin. Wenn Sie es in der IT nicht glauben, okay - aber Sie verstehen schon, dass Marketingschreiber bedroht sind, weil es kaum noch Menschengeist braucht, um schnelllebige Gebrauchstexte für den Tag zu liefern?
Wir sollten uns nicht verschließen, auch wenn es an manchen Stellen schmerzt. Wir müssen die derzeitigen Politiken und Versprechungen unterlassen, die eine Zukunft ohne Geburtsweh anstreben ("Klimawandel muss nichts kosten!"). Wir brauchen auf allen Ebenen eine entschlossene Führung in eine neue Kultur. Wir selbst sollten uns für diese öffnen: "Baue Schiffe, keine Deiche." - "Wer nicht will, wird irgendwann müssen, aber dann kann er nicht, weil er abgehängt worden ist."

Great engineers often use back-of-the-envelope calculations to estimate resources and costs. This practice is equally beneficial in Machine Learning Engineering, aiding in confirming the feasibility and value of an ML project. In my talk, I'll introduce a collaborative design toolkit for ML projects. It includes Machine Learning Canvas and MLOps Stack Canvas to identify ML use cases and perform initial prototyping, thus ensuring a business problem can be effectively solved within reasonable cost and resource parameters.

Target Audience: Architects, Developers, Project Leader, Data Scientist
Prerequisites: Basic knowledge in machine learning
Level: Advanced

In this session, we focus on the topic of software product management (PdM) and how PdM practices are rapidly changing. Together we explore and define how to do PdM for digital products as well as software-, data- and AI-intensive systems. Some questions we explore include:

1. How to change current PdM practices to work with digital technologies and digital offerings?
2. What is the future of PdM practices and what are the key characteristics of digital product management?

Target Audience: Product Managers, Software-Architects, Software Engineers, R&D Management
Prerequisites: No specific technology expertise required.
Level: Advanced