AI/AI Safety Forecasting Workshop + DeepSeek r1 Q and A

​This is an interactive 2-hour workshop that is organized with the AI Club at University of Minnesota, free and open to the public, focused on AI forecasting and safety awareness along with a look at the technical details and wider implications of DeepSeek's newest r1 model.

​Our goal is to help participants understand the landscape of AI and AI safety through interactive demonstrations and group exercises aimed at helping people forecast what might be coming over the next year or two with AI and AI safety.

​The first portion of this workshop will be a set of virtual talks and demonstrations of current AI capabilities.

​The second portion of this workshop will be more detailed forecasting exercises, where we’ll use prediction markets to calibrate expectations and help forecast what kinds of AI developments might come and when.

​Finally there is a final section where we will have a Q and A session about the DeepSeek r1 paper and salient details about the model for those who are more technically inclined. We will go over the r1 paper on arXiv and explain various bits of its terminology, what makes it different from pre-existing models, and what this reveals about OpenAI's own o-X family of models.