80,000 Hours Podcast cover image

#184 – Zvi Mowshowitz on sleeping on sleeper agents, and the biggest AI updates since ChatGPT

80,000 Hours Podcast

NOTE

Challenges of Transparency in AI Development at Google and DeepMind

OpenAI and Anthropic adopt transparency to reassure the public about their AI development, while Google has constraints due to its size and internal processes. The barriers within Google, like involving publicity teams, communications teams, lawyers, and regulatory concerns, hinder the level of transparency seen in smaller entities. This lack of transparency raises concerns as the technology progresses and potentially becomes more powerful, leading to a state of ambiguity for outsiders regarding Google's AI development. The limited visibility into Google's activities may prevent insight into whether they are effectively managing AI risks, leaving outsiders uncertain about the decisions and initiatives within DeepMind and Google.

00:00
Transcript
Play full episode

Remember Everything You Learn from Podcasts

Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
App store bannerPlay store banner