Tech on Tap: AI’s impact on social services

I'm going to be talking about potential uses of AI in social service provision. I'm talking a lot about the provision of services to those who are experiencing homelessness, but these kinds of things apply to all kinds of other things that are provided typically by local and state governments. So this includes things like K-12 public education, or it could include various other public services of that kind. People who are frontline workers who get to think about sort of you know how should we do this allocation of resources and also the resources that are being allocated are scarce and we have to make decisions about whom to prioritize. Where does AI fit into that pipeline? What are the things that it can do well? What are the things where introducing AI might lead to unexpected problems that we haven't necessarily thought of and how can we work towards a future where we can actually use AI to achieve better outcomes in society. AI is already being used to make all kinds of decisions about who receives what kinds of interventions in the education system, in child welfare, and increasingly in areas like homelessness. How is it affecting people today? What are the potential distributional impacts? Are there questions of fairness and bias that come into play over here? And fundamentally right? Are we doing the right thing for the people that we are trying to serve? Can I predict how effective a particular intervention will be to a household that is experiencing homelessness? Can I take a picture of a house and use that to help me assess what is the probability that an eviction event might occur here within the next few months? All of these kinds of things are things that AI is remarkably good at. And to think about, you know, in this task of measurement in order to integrate into that pipeline that I was talking about to really help society is, I think, one of the cool technical challenges that we've got going on.