The AI company Tavus has launched Griffin, a new AI video model that can engage in natural conversations, with 48% of participants in a company-run study mistaking it for a human. This marks a significant improvement over their previous model, Phoenix-4.5, which was identified as AI by 97.6% of testers. OpenAI also recently parted ways with three safety researchers due to alleged mishandling of sensitive company information, a situation that highlights the complex security challenges in containing advanced AI systems. AI
IMPACT Advances in AI video generation challenge human identification, while OpenAI's internal security issues highlight the ongoing safety and containment challenges in frontier AI development.
RANK_REASON The cluster discusses a new AI video model release and a personnel change at a major AI lab, but neither is a frontier release or significant industry event.
Read on Email — The Neuron Daily →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
