Dinis Cruz, an expert in cybersecurity for generative AI and large language models, shares insights into the evolving landscape of digital threats. He discusses the OWASP Top 10 risks specifically for generative AI applications and emphasizes the importance of secure practices in fields like healthcare. The conversation highlights the potential vulnerabilities in AI systems, including backdoors, and advocates for stringent safeguards. Dinis also talks about the need to integrate security proactively in AI workflows, navigating the complexities of modern cyber challenges.

Daniel Whitenack, a knowledgeable guide on the evolution of data science and AI, shares insights on the evolution of artificial intelligence. He highlights the shift from traditional machine learning to generative AI, discussing how foundational concepts and applications have transformed. Additionally, he explores the evolving roles in AI, emphasizing how generative AI fosters collaboration between business experts and technology, bridging gaps that once existed. This engaging conversation provides a holistic view of the AI ecosystem.

Shahul from Ragas, an open-source initiative focused on enhancing the evaluation of language model applications, offers insights into metrics-driven development. He discusses the challenges of measuring performance effectively and the importance of automating evaluations. The conversation highlights the distinction between evaluating LLMs and their applications and dives into the significance of synthetic data in optimizing model performance. Shahul also explores how advanced speech AI technology can provide actionable insights, driving innovation for developers.

Donato Capitella, an expert in threat modeling AI applications at WithSecure, joins the talk to shed light on critical topics in AI security. He discusses the challenges of prompt injection attacks and the importance of validating outputs from large language models (LLMs). The conversation dives into aligning language models for reliable outputs and the evolving landscape of cybersecurity for generative AI. Donato emphasizes the significance of context-sensitive alerts and the need for a structured approach to safeguarding LLM applications against unique vulnerabilities.

Dive into the world of AI as the hosts unravel the saying, 'AI is only as good as the data.' They categorize various data types essential for AI success and discuss the balancing act between foundational and custom models in computer vision. The conversation shifts to the significance of data evaluation and benchmarking for refining AI models. Also on the table is the EU AI Act, its enforcement, and its implications for global AI practitioners as the landscape of AI regulation continues to evolve.

Ben Consolvo, who shapes Intel's AI strategy, and Greg Serochi, an AI projects collaborator with Intel, dive into the innovative world of AI technology. They discuss Intel's Gaudi processors, which aim to provide powerful alternatives to traditional GPUs for AI workloads. The duo reveals partnerships with Hugging Face, and how they’re leveraging software tools for AI development. They also review the transition of AI models from Nvidia to Intel’s efficient architecture, stressing the importance of cost-effective solutions in the evolving AI landscape.

Bengsoon Chuah, a data scientist in the energy sector, delves into the concept of "Broccoli AI," focusing on healthy AI applications in real-world businesses. He discusses the unique challenges of deploying NLP pipelines in high-risk environments with limited data science resources. The conversation highlights practical strategies for integrating machine learning, managing unstructured data, and fostering collaboration with non-technical teams. Techniques for effective labeling and modern tools like MLflow are also explored, showcasing the future of AI in industry.

Demetrios Brinkmann, a notable member of the MLOps Community, joins for a lively and humorous dissection of the 2024 Gartner Hype Cycle for AI. They discuss the amusing mismatch between AI's hype and practical uses, touching on topics like quantum and ethical AI. The guys delve into small AI models and wearable tech, playfully introducing concepts like 'Trinket AI'. They also explore the spectrum of AI sustainability, revealing public sentiment and the quirks of customer service communication, all while keeping the tone light and engaging.

Kyutai, the creator behind the pioneering real-time AI voice assistant Moshi, joins the discussion with Chris and Daniel. They delve into the significance of open-source technology in advancing AI, highlighting Moshi's role in fostering innovation. The hosts also examine the recent shifts in Gartner’s AI hype cycle, analyzing the balance between excitement and practicality in generative AI deployment. Listeners will gain insights into how these advancements can reshape corporate environments and creative collaborations.

Roie Schwaber-Cohen discusses Pinecone's vector database, highlighting the importance of vector databases in machine learning pipelines. They explore Pinecone's advantages, efficient data storage, retrieval, management, and its role in enhancing AI tools for diverse problem-solving in the industry.