The AI4ASE '24 workshop was co-organized with the XP 2024 conference in Bolzano. The workshop took place on June 4th from 14:00 to 17:30. Please see the XP conference website for details about the location.
Start Segment Item
14:00 Introduction Opening and welcome
14:11 Introduction Introductory talk by Clare Dillon, PhD Student University of Galway and Community Lead, CURIOSS
14:36 Research presentations Autonomous Agents in Software Development: A Vision Paper (Zeeshan Rasheed and Sami Malik)
14:54 Research presentations Generative AI for Test Driven Development: preliminary results (Moritz Mock)
15:12 Research presentations Responsible AI in Agile Software Engineering - An Industry Perspective (Rasmus Ulfsnes)
15:30 *** AFTERNOON BREAK ***
16:00 Research presentations A journey through SPACE: Unpacking the Perceived Productivity of GitHub Copilot (Viggo Tellefsen Wivestad)
16:18 Research agenda Introduction
16:28 Research agenda Interactive session - all involved
17:18 Closing (ending 17:30)
The audience was invited to share their ideas on research questions and topics. The transcribed notes are listed below and grouped according to the WS organizers’ interpretations.
Notes (grouped)
Quality control, trustworthiness, and ethics
How can we recognize AI-generated output?
How can we tell what code is AI vs human generated?
How to perform quality assurance on AI answers?
What areas of SE can be (safely) automated?
Ethical considerations
AI + semantic models; how can we generate [?] “verifiable” (formally) AI output?
How can AI-results be testable?
Right orchestration and cost-effective QA of AI tools for your use case
Humans and AI
How do humans and AI collaborate?
How can developer satisfaction be maintained?
What happens to team collaboration when everyone has their own AI-assistant?
Agentic workflows
Learning and adoption
How to train organizations to leverage AI responsibility?
How is AI going to affect software engineering education? Will AI negatively impact the learning outcomes?
What are the key factors for adoption/implementation of AI?
Secondary school education is also already impacted by AI, how do we need to change higher school education?
What is the most efficient way to build AI-skills quickly?
Usefulness and effects of AI
Are there measurable quality differences in AI vs human-generated code?
Long-term effects of AI-based coding assistants on learning and team communication
Does AI fit for full product development, or MVP and experiments?
Where should product owners start to automate software engineering with generative AI?
What is the impact of new architecture beyond GPTs? [unclear note]
How does AI influence agile SW development?
Can AI be used for coordination in agile dev?
How will AI-agents change SE?
How to ensure knowledge organization beyond/despite of AI?
Can AI help to develop better software products?
Practical guidelines
Practical guidelines for the use of AI in software engineering
What are the best practices for GenAI/LLM usage, what are the problems that LLMs can solve?
What do we actually need of AI in SE?
AI x knowledge generation. Where/how can AI help generate new knowledge about software?
AI x Software development process. How and where can we integrate AI in the software development process?
Further research and theory building
How to get research to practitioners faster?
Which existing or new theories exist for AI integration?
What comes after GPT & LLMs?
How to not get locked in with OpenAI?
Leveraging XR and AI and multimodal capabilities
Astri Barbala, SINTEF
Viktoria Stray, University of Oslo
Geir K. Hanssen, SINTEF
Klaas-Jan Stol, Univ. College Cork
Jan Henrik Gundelsby, Knowit
All published workshop papers should be short papers. Camera-ready copies of revised papers along with a signed license to publish documents should be submitted by Wednesday, 26th June. (More details will follow).
Licence agreement (will be shared later)