The conversation centers on the themes of AI agents, reliability, and H-1B visas, particularly in light of discussions at the recent AI Conference. Evangelos Simoudis joins Ben Lorica to discuss the gap between the promotion of AI agents and their actual implementation in enterprises, noting high costs and cybersecurity risks associated with "shadow AI." They emphasize the need for solid data infrastructure and workflow rethinking before deploying AI solutions. Shifting to reliability, they address challenges like hallucinations and non-deterministic outputs, advocating for rigorous software engineering practices and real-time data integration. Finally, they critique the new H-1B visa policies, arguing that the $100k salary threshold could hinder startups and small companies from accessing crucial talent, potentially leading to a brain drain and unfilled manufacturing jobs.
Sign in to continue reading, translating and more.
Continue