AI in Power BI and context

0:00 I'm talking about AI actually creating measures, setting up relationships, actually doing development work because that is possible now since the release of MCP servers or PowerBI which basically lets you plug in your semantic model to a large language model to AI. Now that is a game changer. However, nobody is addressing the big elephant in the room. Is it safe? Should we have some data privacy concept? Well, let's dive in a little bit deeper here. When chat GPT was released at the end of 22, it caused a big revolution because people like you and me suddenly had access to AI and well, I started using it for everything.

0:49 I still do. But I started small, for example, with having it write my birthday card messages because well, I never know what to write. However, then I also moved on to bigger things. For example, business emails, having it check legal documents or having it help me write DAX measures. Well, not that I need it, but you know. The thing is though, the answers that I would get would often be kind of generic because the model that you're using, well, it is trained on large amounts of text that it's scraped from the internet and it just knows that.

1:12 It doesn't know anything about your PowerBI model. So therefore it can help you just a bit kind of in a generic way. But if you give it more contextual information, the answers you would get would be better. For example, if it would know the column names and all of your tables, the granularity of your data, well then it can write much better DAX. Now another thing is that well AI would not be able to directly do the work for you like write your DAX measures and that's where MCP servers come in.

1:35 Now let me give you an overview of how these MCP servers are going to change your PowerBI development work with AI. Now let's say we got you