AI Today Podcast

The podcast discusses the highlights of AI in 2023, including generative AI and Trustworthy AI. Venture Capital investment in AI remains strong. Topics covered include recent investments, protein folding breakthrough, advancements in large language models, trends and developments in AI, and upcoming topics.

The podcast discusses the groundbreaking EU AI Act, the world's first comprehensive AI law, which aims to regulate AI use in the EU. It explores the different levels of AI systems regulation and the implications of the Act for businesses, including compliance requirements and regulation of image manipulation tools. The episode also emphasizes the importance of a trustworthy AI framework and provides resources for further learning.

Nandini Nandakumar, a marketing expert at Diageo, discusses the role of AI in marketing and product innovation. They explore predictive modeling, consumer segmentation, and content creation. The podcast highlights the importance of collaboration, starting small, and addressing challenges in AI adoption. It also discusses the future of AI in the Consumer Packaged Goods industry.

The hosts discuss the different lengths of time to see ROI for various AI projects, including augmented intelligence and assisted intelligence. They explore the concept of using augmented, assisted, and autonomous systems in AI projects to achieve a faster ROI. The varying timeframes to achieve ROI for different types of AI projects are highlighted, such as conversational systems having a shorter time to ROI compared to predictive analytics and autonomous activities.

The AI Today podcast explores when to use automation versus AI, including examples of automating repetitive tasks and the suitability of AI for complex tasks. It also discusses scenarios where automation falls short, such as image recognition and customer service. Additionally, the podcast highlights the strengths and weaknesses of both approaches and introduces the CPMAI methodology and certification.

Xiong (Sean) Liu, Director at Novartis, discusses the impact of AI in pharma including streamlining processes, optimizing drug discovery, and enhancing patient care. He shares real-world examples of AI applications in biology discovery, molecule design, and clinical trials. The chapter also emphasizes the transformative impact of AI and the importance of ethical AI in the pharmaceutical industry.

The podcast discusses the difference between proof-of-concept, pilot, and production projects in AI. It emphasizes the importance of pilot projects in testing real-world scenarios and highlights the need to avoid assuming that a proof of concept will always lead to success. The hosts also discuss the CPMAI methodology and its benefits for AI project management.

Hosts discuss important AI terms like anonymization, GDPR, and the uncanny valley. They emphasize the value of dialogue with the audience and the need for data anonymization. The podcast explores the impact of GDPR on AI systems and the balance between personalization and privacy. It also delves into individual comfort levels with privacy and the role of the uncanny valley in AI system failures.

The hosts discuss the glossary series of the podcast, delve into the concepts of Hadoop and MapReduce, and highlight the importance of understanding unstructured data and how machine learning can assist in its analysis. They also explore the challenges of managing and analyzing large amounts of unstructured data, introduce the concept of map reduce, and discuss the features and components of Apache Hadoop. Additionally, they explore the CPMAI methodology for running AI projects and the importance of understanding and utilizing terms related to AI.

In this podcast, the hosts discuss algorithmic discrimination, governance, and pseudo AI. They explain the consequences of biased data and lack of transparency, and emphasize the importance of human involvement in addressing bias. They also explore the role of governance in AI systems and the issue of companies using humans instead of AI without disclosure.