Mars Missions: Past, Present, and Future. From fiction to Reality

Abstract

In this talk we will explore our long-standing curiosity and fascination with the Red Planet, tracing the evolution of Mars exploration from early science fiction to cutting edge missions. We will cover significant milestones in Mars exploration, and the natural progression of the various missions that attempted to reach or to land on Mars. This would span from NASA's Viking landers, Spirit and Opportunity rovers, to the recent achievements of Perseverance and Ingenuity. We will discuss upcoming missions, such as the Mars Sample Return, along with the bold visions and plans for landing humans on Mars. We will explore the scientific and technological challenges of reaching and eventually inhabiting Mars.

The Frontier of Intelligence: How Generative AI and LLMs Are Reshaping the Business Transformation

Abstract

In this talk, I will explore the remarkable evolution of large language models (LLMs) since the inception of groundbreaking technologies like ChatGPT. I explore the entire lifecycle of these models—from their complex pre-training phases through the nuances of instruction fine-tuning to the sophisticated realm of Reinforcement Learning from Human Feedback (RLHF). This discussion will delve into the advanced methodologies that shape these models, focusing mainly on Agentic LLMs and the dynamic arena of multi-agent collaboration frameworks.

As industries worldwide seek innovative solutions to complex challenges, LLMs stand at the forefront of this transformative era. We will showcase a series of compelling business use cases where LLMs have not just participated but significantly revolutionized fields such as healthcare, education, and research. From advanced chatbots that redefine user interaction to AI-driven platforms that enhance medical diagnostics and personalized learning experiences, the influence of LLMs in reshaping the business landscape is undeniable.

This talk aims to inform and spark thought on leveraging the power of LLMs to drive forward the most advanced and complex business transformations.

Smart Systems and Emerging Technologies for Smart Cities: Leveraging Digital Twins for Pollution and Traffic Assessment

Abstract

As urban areas continue to expand, the need for efficient and sustainable management of pollution and traffic becomes increasingly critical. This keynote explores the transformative potential of smart systems and emerging technologies, specifically focusing on the application of Digital Twins in smart cities. Through the integration of IoT sensors, data analytics, and AI-driven models, Digital Twins offer a comprehensive approach to monitoring and mitigating urban pollution and traffic congestion.

Drawing on the experiences and success stories from Libelium's extensive deployments in over 100 cities worldwide, this presentation will highlight how Digital Twins can provide real-time insights and predictive capabilities. These systems enable urban planners to make informed decisions, optimize traffic flows, and implement effective pollution control measures. Case studies will demonstrate the practical implementation and benefits of these technologies, including reduced CO2 emissions, improved air quality, and enhanced urban mobility.

Attendees will gain a deep understanding of the key components and functionalities of Digital Twins, the importance of data standardization and integration, and the role of AI in forecasting and scenario analysis. The session will also address the challenges and opportunities in deploying these technologies at scale, and how they contribute to the broader goals of sustainability and smart city development.

Multimodal Arabic Large Language Models

Abstract

Arabic-centric large language models (LLMs) face numerous challenges, including inadequate evaluation frameworks, cultural insensitivity, limited representation of diverse Arabic dialects, lack of multimodal capabilities, overly generic designs for specialized domains, and a disconnect from other low-resource languages. In this talk, we address several of these challenges by developing models capable of understanding and generating content across a broad range of Arabic languages and dialects. We introduce a suite of LLMs specifically tailored for text, speech, and image, designed to enhance Arabic representation in key domains. By focusing on areas such as archival work, cultural heritage preservation, financial services, healthcare, and education, our work aims to bridge the linguistic digital divide and ensure equitable access to AI. We will discuss our methods, the challenges encountered, the solutions proposed, and the broader implications of our efforts.