Nathaniel Whittemore reviews proposals for public participation in AI company equity and contrasting views about their implications. He also discusses reported compute dealsAI compute infrastructure is the hardware, facilities, networks, power, cooling, storage, and software used to train and run AI models. and supply-chain constraints.
Nathaniel Whittemore then considers reporting about changes to ChatGPT and argues that the move toward agents reflects a gap between casual and advanced AI use. The plans and motivations are presented as reporting and interpretation, not settled outcomes.
Nathaniel Whittemore describes agent loopsAn agent loop repeatedly observes a task's state, chooses an action, checks the result and decides whether to continue or stop. as a way to automate repeated prompting and notes that access, instruction and error accumulation remain challenges.
Watch on YouTube
