Marina Danilevsky

This discussion features Gabe Goodhart, an AI architect focused on cybersecurity, Merve Unuvar, a specialist in agent middleware, and Marina Danilevsky, a research scientist analyzing AI model behaviors. They dive into Google’s Gemini 3 model, exploring its strong performance yet concerning hallucination issues. The dialogue shifts to AI's impact on the economy through OpenAI’s GDPVal benchmark, and the panel debates the balance between specialized and generalist models. They also tackle the implications of a recent cyberattack automated by AI, stressing the need for robust enterprise defenses.

Join experts Marina Danilevsky, an innovative senior research scientist, Nathalie Baracaldo, a master inventor, and Sandi Besen, a seasoned AI research engineer, as they unpack the fallout of a report revealing that 95% of generative AI pilots miss enterprise targets. They dive into GPT-5's hidden prompts, exploring transparency challenges in AI development. The conversation shifts to the shortcomings of reasoning models and the ethical implications of Claude's decision to end distressing conversations, raising important questions about AI welfare.

Marina Danilevsky, research scientist focused on practical, safe community-driven agents. Kaoutar El Maghraoui, principal research scientist specializing in AI platforms and system strategy. Aaron Baughman, IBM Fellow and inventor working on fan-facing AI systems. They discuss Moltbot’s open-source surge and Mac mini craze. They debate agent usability versus benchmarks, agent security and sandboxing, and chip verticalization with Microsoft’s Maia 200.

Join Marina Danilevsky, a senior research scientist in AI, Abraham Daniels, a technical product manager, and Gabe Goodhart, chief architect in open innovation as they unpack the latest from Google I/O 2025. They delve into whether Google’s AI Ultra subscription is worth it and discuss the implications of RedHat's new llm-d for Kubernetes. They also tackle Microsoft’s NLWeb and the transforming nature of search. Finally, the trio examines Stack Overflow's decline, pondering if AI is pushing it towards obsolescence.

Kush Varshney, an AI expert from IBM, joins Martin Keen, a Master Inventor, and Marina Danilevsky, a Senior Research Scientist, to discuss Disney's bold $1 billion licensing deal with OpenAI. They delve into how this partnership aims to control fan-generated content and enhance Disney Plus traffic. The panel also evaluates the implications for IP owners and debates the evolving dynamics of authorship. Additionally, they analyze the significance of Time Magazine's recognition of AI architects and unpack NVIDIA's new model releases.