Blog

World AI Week 2025: from intelligent systems to neuro-inspired futures

Amsterdam, 6-10 October 2025 —The latest edition of World AI Week once again turned the Dutch capital into a global meeting point for artificial intelligence. With more than 7,000 researchers, innovators, and policymakers gathered across dozens of sessions, the event reaffirmed AI’s growing role as an infrastructure for science, industry, and society.

From the World Summit AI to side events on AI for Climate, AI in Creative Industries, and AI Governance, the agenda covered the full spectrum of current challenges, from algorithmic transparency to compute scalability. Several themes stood out:

• Autonomous and embodied systems, advancing robotics, sensorimotor learning, and adaptive agents.
• Scalable AI infrastructure, including neuromorphic and quantum approaches to efficient computation.
• Responsible and ethical AI, shaped by the EU AI Act.
• Applied AI across healthcare, sustainability, and industrial automation.
• Human-AI collaboration, highlighting low-code development and creative co-design.

While much of the week focused on strategy and cross-sector dialogue, a few sessions brought genuine technical depth, most notably the NVIDIA workshop on extending large language models with structured knowledge.

NVIDIA’s “Adding Knowledge to LLMs” Workshop

Held on 7 October at NXT Lab, the “Adding Knowledge to LLMs” workshop offered participants a hands-on exploration of methods to augment large language models with external knowledge graphs and domain-specific datasets. The session reflected a shift in the AI community: progress is increasingly defined not by scaling parameters, but by integrating structure, efficiency, and reasoning into existing systems. As industrial and research teams move toward knowledge-aware architectures, the frontier of AI appears to be moving from “bigger” to “smarter.”

Gender Balance and Leadership in AI

Another visible theme throughout the week was the continued emphasis on gender balance and representation in the AI ecosystem. Several community events showcased women’s leadership in technical and governance roles:

• “Leadership of Strategic AI Initiatives” by Women in Product NL, exploring pathways for women leading large-scale AI projects.
• “Women Leading Tech – Safeguards of AI for Good”, a public stage highlighting ethical and human-centric leadership.
• “Women Building AI: Beyond Boundaries”, hosted by Women in AI Benelux, celebrating engineers, researchers, and entrepreneurs shaping responsible innovation.

The event organisers, led by Sarah Porter, founder of InspiredMinds, have made inclusion a defining principle since the inception of World Summit AI. Her approach demonstrates how community design can be used to correct structural imbalances in the field.

This year’s programme also featured several women in technical leadership roles, including: Helia Mohammadi, Chief Research Technology Officer at City of Hope, who bridges biomedical research and AI engineering. Jessica Dunn, Global Technology Officer – Applied Innovation at EY, leading the translation of AI research into scalable industrial solutions. And Payal Arora, Professor of Inclusive AI Cultures at Utrecht University, whose work connects technical design with cultural and social impact. Their contributions illustrate that equitable progress in AI is inseparable from technical excellence and inclusive leadership.

Looking Ahead: The Next Frontier

Although the 2025 edition of World AI Week largely emphasised strategic and societal dimensions, certain sessions, from NVIDIA’s technical deep dives to references to neuro-adaptive systems and discussions on adaptive AI, hinted at a deeper scientific convergence on the horizon. That intersection where computational architectures meet biological insight, which defines the emerging field of Neuro-AI.

Neuro-AI seeks to bridge neuroscience and artificial intelligence, taking inspiration from how real neurons and networks compute, learn, and adapt. From spiking neural networks and neuromorphic hardware to learning rules grounded in biological plasticity, this approach aims to build systems that are not only powerful but also efficient, interpretable, and resilient.

For the MERIT project, this convergence represents an opportunity to connect disciplines, from computational modelling to cognitive systems and responsible innovation, as we collectively imagine intelligence beyond algorithms, moving towards architectures that learn like brains and collaborate like societies.