Exploring the Future of AI: Key Moments from the Ai4 Global Conference

According to Dr Agnė Paulauskaitė-Tarasevičienė, professor at Kaunas University of Technology (KTU), attendees had the opportunity to network with key decision-makers, researchers and developers from leading companies and organisations. They explored potential partnerships, joint research and development activities and workshops to advance Centre of Excellence of Artificial Intelligence for Sustainable Living and Working (SustAInLivWork).

“It was very important to gain insights into effective strategies for implementing AI solutions, to see real examples of successful AI projects (especially in the USA) and to learn from their experiences. In addition, it was crucial to understand the current and future landscape of the AI industry and to identify new technologies that could be relevant for planning the future services of CoE SustAinLivWork,” notes Prof. Dr Paulauskaitė-Tarasevičienė.

Additionally, the event facilitated discussions with academic experts from institutions such as the University of Kansas, the University of Waukesha, the University of Arizona Global Campus and the Institute for Responsible Artificial Intelligence, focusing on potential joint projects in health and manufacturing, two of the four main CoE sectors.

“This event had a special focus on networking among participants. We took the opportunity to connect with other higher education institutions and companies involved in the development and application of AI solutions. We plan to actively utilise these contacts in the SustAInLivWork project, which is establishing a new AI Competence Centre and will offer a range of AI services. It has been an excellent opportunity to deepen our knowledge, expand our network, and prepare for future challenges in developing innovative AI solutions,” says Dr Simona Ramanauskaitė, professor at Vilnius Gediminas Technical University (VILNIUS TECH).

Applications of AI

The conference featured more than 100 presentations covering current issues and the future of AI. For example, it explored how social order, politics, and other social aspects might change when AI is applied when making decisions that affect society.

“I enjoyed the topics related to the application of AI to medical data analysis. The speaker, David Lederman, who discussed the application of the generative AI algorithm in elderly care, attracted a lot of attention. His presentation demonstrated examples of how these technologies contribute to improving the health of seniors,” shares Dr Vidas Raudonis, professor at KTU.

As highlighted by Prof. Dr Raudonis, a particularly intriguing topic at the conference was the exploration of decision-making mechanisms within deep neural networks. One of the standout methods discussed was the identification of the most significant inputs influencing a neural network’s final decisions. By ranking these inputs according to their importance, this disaggregated model offers valuable insights, particularly in medical contexts. This approach can help explain to patients the underlying reasons for specific medical procedures, fostering trust and understanding.

“Attending Ai4 significantly broadened my understanding of AI’s potential applications,” stated Prof. Dr Raudonis. “From analysing MRI images to diagnose cancerous lesions to improving the mobility of the elderly using generative AI, AI is revolutionizing various fields. Even in the realm of politics, AI is being explored as a tool to assist in decision-making.”

Strong consensus was found

Speakers and participants shared not only their experiences of how certain AI-based solutions are addressing today’s challenges but also their concerns. For example, a former Democratic candidate for the U.S. presidency expressed worries about the potential threat of AI to jobs and its impact on social security. There was a common perception that, currently, AI automates cognitive rather than biomechanical human functions, so for the time being, long-distance drivers can feel relatively secure about their jobs.

Prof. Dr Raudonis notes that research is increasingly focusing on more transparent AI models, commonly known as Explainable AI (XAI) models. He highlights a strong emphasis on generative models capable of producing diverse visual content. Furthermore, a substantial portion of the discussions focused on large language models (LLM) and their promising future applications.

The participants that representatives from the SustAInLivWork project engaged with at the conference expressed a common mission and strategy for the centre, in line with the trajectory of AI development in the Baltics. A strong consensus on the most promising areas for AI innovation in our region was found.