Tag
Addy Osmani reflects on the role of taste and judgment in an AI-driven world, arguing that while AI can acquire taste, human judgment—rooted in accountability and ownership—remains irreplaceable.
Explores why humanity created digital intelligence that it does not fully understand, delving into philosophical and technological implications.
This paper studies a human-AI service system with an automated chatbot and human agents, proposing a UCB-DPP policy that learns unknown parameters and achieves regret Õ(K√T) while stabilizing queues.
This paper introduces a toy framework that models curiosity as an ecosystem in single and multi-agent settings, exploring how agents weigh immediate uncertainty reduction, costs, delayed returns, and the value of keeping questions open. It aims to inform future multi-agent AI systems for discovery.
Jerry Liu agrees that both agents and software have value, but notes that their interfaces are different—agents use simple communication interfaces like chat, while software tools need tailored interfaces for specific tasks.
A new process-level latent variable model (PLVM) predicts future behavioral strategies from partial process traces across tasks, demonstrated in PowerWash Simulator gameplay data.