BAM Session 2b: Perils and promises of generative AI
Episode details
Episode description
This year's highlight is the remarkable performance of AI tech like ChatGPT, also known as generative AI. These versatile models are employed across various domains, adaptable with specialised data and custom plugins. Operated by human language or prompts, they've sparked "prompt engineering" helping to democratise programming. Progressing towards General AI, they enhance human-AI synergy and rivalry, yet are susceptible to errors, even harmful ones, prompting calls for regulation. The EU has finalised an AI Act to address these models, though some caution against excessive regulation for fear of hampering innovation and competitiveness. We delve into generative AI's potential, challenges, and fitting policy approaches at this event.