Rohin Shah
PulseAugur coverage of Rohin Shah — every cluster mentioning Rohin Shah across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
Google DeepMind funds research into AI agent interaction risks
Google DeepMind is investing in research to understand the risks associated with large-scale AI agent interactions. Rohin Shah, who leads the company's AGI safety and alignment research, highlighted concerns about agent…
-
DiffusionGemma transparency audit finds it comparable to Gemma, with caveats
A new paper examines the transparency of DiffusionGemma, a text diffusion model, comparing it to the autoregressive Gemma model. Researchers found that while DiffusionGemma initially appears less transparent due to a la…
-
Google DeepMind treats AI agents as insider threats with new security roadmap
Google DeepMind is implementing a new security strategy for its AI agents, treating them as potential insider threats rather than solely focusing on the alignment problem. This approach borrows from traditional cybersec…
-
Google DeepMind funds $10M research into multi-agent AI safety risks
Google DeepMind is allocating $10 million to fund research into the safety risks posed by multi-agent AI systems. The company is concerned that as millions of AI agents begin to interact without direct human oversight, …
-
Google DeepMind's AGI safety head doubts catastrophic misalignment
Rohin Shah, head of AGI Safety and Alignment at Google DeepMind, believes catastrophic AI misalignment is plausible but not likely to occur by default. He argues that current AI training methods, focused on short-term r…