Project summary

AI Safety & Society (AISS) is a new platform dedicated to enabling high-quality and timely discourse on AI safety. The platform will publish articles by prominent intellectuals appealing to a wide audience by discussing various facets of AI safety. The platform would occupy a critical and currently nonexistent niche between rapid, informal social media discussions and formal, but slower academic publications. By providing a space for accessible, in-depth, rigorous, and rapid discussions on emerging issues, the platform aims to elevate the quality of public discourse on AI safety and inform a broad audience on the field’s most important developments and ideas.

What are this project's goals? How will you achieve them?

AISS’s primary goal is to create and sustain a platform akin to other resources like Lawfare and to build this platform into a nexus for AI safety discussion and learning. We will achieve these goals by maintaining a predictable publishing cycle, by constantly acquiring new articles from our broad network of academics, business leaders, think tank fellows, and other industry professionals. We will retain in-house capacity in writing and editing to ensure a smooth operation and effective quality control. And finally, we will carefully vet our writers to ensure their articles are relevant and coherent, while making sure not to otherwise limit or censor any ideas.

How will this funding be used?

We will use the funding provided to cover staff salaries (writers, an editor, a PM) and market our platform. An editor is needed ensure that all articles meet a certain standard of quality and rigor, and are free of unnecessary jargon or clutter. The editor would also take a lead role in deciding when each article comes out. A project manager would identify prospective, extenral writers, reach out to them, and generally oversee the success of the project. In-house writers would produce a steady stream of content for the platform. And lastly, we would spend money on marketing the platform to ensure that it has the maximum possible reach.

Who is on your team? What's your track record on similar projects?

Dan Hendrycks (Executive Director at CAIS), Oliver Zhang (Managing Director at CAIS), Richard Crook (Communications Director), are currently on the AISS team. Dan received a Ph.D. from UC UC Berkeley. He has an h-index of 38 and his papers have been cited over 32,000 times. Prior to CAIS, Oliver co-founded ML Alignment Theory Scholars (MATS). Before joining CAIS, Richard spent over a decade at the BBC, where he served as a senior journalist and editor before becoming a special advisor to the Chairman.

What are the most likely causes and outcomes if this project fails?

1. Audiences. Growth in a very live and competitive marketplace of ideas will be challenging. There are many places people can find news and analysis about AI so we will need to carve a USP that makes us distinct 

2. Editorial depth, breadth and quality. This is a relatively small field, and we need a large cohort of talented writers, op-ed contributors and in-house editors who know the field well and can be engaging and accessible, while delivering genuine expertise. And doing this daily and weekly. 

3. Marketing - Starting a website from scratch outside existing blog infrastructure and building audiences requires a multifaceted plan to spread the word.

How much money have you raised in the last 12 months, and from where?

None