About us
To all people who want to change the world using new technology and who get a kick from hearing words like predictive modelling, natural language processing, and statistical inference, come and share your machine learning experience or interests. We will have presentations and workshops from both academia and industry.
The non-profit association Gothenburg Artificial Intelligence Alliance hosts this group. The organisation also hosts a yearly conference, the GAIA Conference. GAIA aims to promote the interest in artificial intelligence, machine learning, data science, and data engineering in Gothenburg and the surrounding area.
You can read more about our organisation on our webpage: https://www.gaia.fish. We have a YouTube channel where we occasionally share event recordings: https://www.youtube.com/@GAIA_GBG.
If you like our events and want to keep in contact with the network between events, check out our Slack at https://www.gaia.fish/get-involved.
Upcoming events
1
GAIA - September Meetup
United Spaces Lindholmen, Pumpgatan 1, 41755, Göteborg, SEAgenda:
17: 30 Welcome from Developers Bay & GAIA
17:45 Carlos Natalino
18:30 Networking & light refreshments 🍕🥤
19:45 Wrap-upThe vision of AI-native 6G networks promises unprecedented levels of automation, adaptability, and intelligence, with AI/ML models embedded at every layer of the network stack. However, realizing this vision requires a fundamental shift in how we assess and deploy AI systems in telecommunications infrastructure. Traditional performance metrics centered on accuracy and prediction quality are insufficient for the complex, safety-critical, and dynamic requirements of next-generation wireless networks. Human intervention inhibits scalability in generating/updating models, and hurts their timeliness in a constantly evolving environment such as 6G networks.
This talk examines the critical role of trustworthy AI in achieving truly AI-native 6G networks and identifies significant gaps that currently hinder this realization.
We discuss two major gaps preventing the full deployment of trustworthy AI in 6G contexts. First, we argue that holistic performance assessment must extend beyond accuracy/error to encompass reliability, robustness, explainability, fairness, and resource efficiency. From the perspective of network automation, we explore why conventional metrics fail to capture the behavioral characteristics essential for autonomous network operations, including model stability under distribution shift, uncertainty quantification, and real-time adaptability to evolving network conditions.
Second, and affected by the first gap, the absence of fully autonomous AI pipelines capable of managing the entire AI/ML lifecycle at the scale with the agility and low human intervention demanded by AI-native architectures. By examining these challenges through the lens of network automation and 6G requirements, we highlight the research and engineering efforts needed to bridge the gap between trustworthy AI principles and operational reality in next-generation networks.
Speaker:
Carlos Natalino is a researcher with Chalmers University of Technology.
He researches trustworthy AI for 5G/6G infrastructures and networks, and their responsible adoption as an enabler for higher levels of autonomy in resilient, efficient, and effective communication networks.Dr. Natalino has been involved in Swedish and European research and innovation projects in close collaboration with industry. He holds two patents. He is an enthusiast of open research, having published data and code of several of his publications.
He is also an active contributor and a member of the Technical Steering Committee of the ETSI TeraFlowSDN Software Development Group.
Welcome!
19 attendees
Past events
60
