Hi! I’m Solana and I’m working with Open Knowledge on the AI Learning Labs.
I don’t know anyone who isn’t feeling pressure to learn more or do more with AI. Many organisations we speak with are hesitant to use tech they don’t fully understand the consequences of using—but at the same time feel it’s unavoidable and necessary. There’s generative AI, of course, and an endless array of productivity tools… And then there are all the intriguing ideas for using AI for mission-related investigations, data analysis, new forms of communication… It’s hard to know where to turn for advice (don’t say ChatGPT!)
I’m wondering whether folks in this forum have recommendations for learning resources or online courses about AI that are especially suitable for people working in social sector organisations. They can be in any language.
This report, New Tech, New Rules by International Resources for Impact & Storytelling (IRIS) has links to a bunch of interesting resources, and 10 interesting case studies of different ways activists and non-profits around the world have used AI in creative new ways.
I’m a scientist/coder so this is probably peripheral…
I’ve been doing a lot of coding supported by Cursor and I find it useful as a tutor. Let’s say I wanted to create a tool to summarize text. I will ask Cursor to review the current approaches and the tools that support each. It will present the pros and cons in a way that I rapidly make an informed choice.
In helping to develop the AI Leaning chatbot system I will ask Cursor to explain the components of the system, what they do and how they interact. In many cases I will ask it to create a Mermaid or similar diagram of workflow, dataflow, etc.
I’m starting from an informed position so I have less experience on how it treats beginners. But remember it (or it’s models) have scraped millions of beginners’ tutorials so it will do very welll here. It won’t be a Feynman but it will be a competent Assistant Professor.
I don’t know how well its database works on social sciences. I’d guess it was good on data collection and statistical analysis. For more general topics it will be very important to iteratively refine the guardrails and this may fail.