\n<\/p>\n
Google and Google DeepMind researchers launched the DeepMind Institute on Wednesday to advance the conversation around artificial general intelligence (AGI). The institute lists DeepMind co-founder Shane Legg, Google executive James Manyika, and Google DeepMind chair Demis Hassabis as directors, with Legg serving as managing editor.\u00a0<\/p>\n
The new institute aims to surface differing views between Google, Google DeepMind, and the broader global research community around AGI. \u201cThey will not always agree, and they will likely change their minds, as more data and information comes to light at the fast-moving frontier,\u201d the announcement read.<\/p>\n
The inaugural collection of four essays covers a range of topics: economic policies for managing potential AGI disruption, preserving human-readable model reasoning, principles for human flourishing, and a framework for evaluating frontier AI models.\u00a0<\/p>\n
One essay<\/a>, by DeepMind safety researchers Rohin Shah and Anca Dragan, argues that AI\u2019s shrinking window of transparency \u2014 the ability to see and check a model\u2019s step-by-step reasoning \u2014 is not inevitable. As new architectures make the most powerful models harder<\/a> to monitor, the authors say developers and regulators should confront the safety trade-offs directly. That could mean limiting \u201copaque serial depth\u201d\u2014 the amount of sequential computation a model can perform without producing a readable reasoning trace \u2014 or requiring developers to demonstrate that less transparent systems remain just as monitorable.<\/p>\n In another essay<\/a>, Hassabis proposes a U.S.-led frontier AI standards body to evaluate the most advanced AI models. Under his framework, developers would initially submit models voluntarily for review up to 30 days before release. Once the evaluation system has proved effective, passing its tests could become a requirement for deploying frontier models in the United States.<\/p>\n The body would at first design assessments in consultation with AI companies but would eventually develop independent, undisclosed evaluations \u2014 what the essay calls \u201cheld-out\u201d tests \u2014 to prevent labs from tailoring their models to known evaluations. Hassabis said the framework could be \u201cratcheted up if the seriousness of the situation demands,\u201d potentially including a coordinated slowdown among frontier AI developers.\u00a0<\/p>\n The essays arrive as the industry\u2019s safety debate shifts<\/a> from broad statements of concern toward concrete proposals for disclosure, outside scrutiny, and, if safeguards fall behind, coordinated slowdowns. That shift accelerated this week as industry leaders endorsed elements of Anthropic CEO Dario Amodei\u2019s call<\/a> to \u201cpace\u201d frontier AI development.<\/p>\n<\/div>\n When you purchase through links in our articles, we may earn a small commission<\/a>. This doesn\u2019t affect our editorial independence.<\/em><\/p>\n