From isolated agents to reasoning together: exploring the collective intelligence of machines
Throughout her PhD and postdoctoral studies, Olga Galinina has been fascinated by the field of machine learning (ML). This interest recently led her to gradually shift focus towards ML solutions in wireless networks, where she aims to combine her mathematical background with engineering and wireless technology perspectives. Now a Senior Research Fellow with Tampere Institute for Advanced Study, she has merged her lifelong passion for research with the potential to make meaningful contributions to the field of collaborative machine learning.
Olga Galinina’s current work involves exploring the field of collective machine intelligence, where different types of machines communicate and interact with each other.
– My goal is to use human-centric collective intelligence and networking solutions to enhance the learning and decision-making abilities of machines, making them more intelligent and creative, Galinina explains.
– I believe that we can create machine systems that can learn and reason collectively without sharing their underlying data or models.
Bridging the gap between different fields to advance AI and its impact on society
The central research challenge is how to design the ability of machines to efficiently exchange information and communication strategies that optimize such solutions in communication networks and drive them to convergent conclusions. Solving this problem calls for multidisciplinary work at the intersection of Communications Engineering, Computer Science, and Cognitive Science, bringing new intertwined problem formulations to the table. This includes all stages of the system, from individual intelligence and reasoning to system-level performance evaluation and optimization.
The demand for sophisticated AI increases
The interdisciplinary nature of Galinina’s research brings new opportunities contributing to the advancement of artificial intelligence (AI) and what effects it will have on society as a whole.
– For example, the recent surging popularity of natural language processing (NLP) models such as GPT-3 and its derivative ChatGPT is absolutely breathtaking and represents a clear indication of the growing demand for more sophisticated AI systems. I believe this intensifies the significance and urgency of research in collective machine intelligence, Galinina says.
– By taking a human-centric perspective, we could not only provide a more intuitive and collaborative approach to machine learning but also create solutions for continuous, lifelong, and agile collaborative learning of heterogeneous machines. Through my research, I hope to lay the foundational blocks in collective reasoning to enable future transitions to scalable and ubiquitous machine intelligence.
Keywords: machine learning, AI, artificial intelligence, Tampere University, Tampere Institute for Advanced Study