We did it again!
Yesterday, as an April Fool, we published an article written by ChatGPT-4 (Revolutionizing Power BI: Introducing the Ultimate Formula Language), which described better than I could have ever done the excitement for a new revolutionizing language: A syntax that introduces hierarchical calculations as a first-class concept in the DAX language.
Clearly, it was a joke, a nerdy joke.
However, what you have seen in the article/video is real: the syntax works, and you can create hierarchical calculations with it.
So, where is the joke?
Well, it’s just not a new thing.
The Ultimate Formula Language we used is just MDX, a syntax that has been used in Microsoft business intelligence products since 1998. It is currently used by Excel when it is connected to a Power BI model.
You cannot create MDX calculations in Power BI; they are not part of the semantic model. However, they can be created in any MDX query and the DAX engine actually also speaks MDX very well!
Hierarchical calculations are a missing part of DAX. The new visual calculations introduce concepts to deal with that, but currently, these concepts are not part of the semantic model and cannot be used in regular DAX measures. Moreover (a joke in the joke), the syntax used by MDX is so much easier than the corresponding DAX! But MDX is so much more difficult than DAX in many other areas – if you think that DAX is hard, you will reframe your way of thinking after learning MDX well.
DAX is powerful, and with great power comes great responsibility, which also requires some learning. Still, a fraction of the learning and practice time you need with MDX.
Unless you want to calculate the ratio-to-parent, which is so easy in MDX.
As you see, it’s April 2, and we are back to almost serious mode. But once a year, we have fun with the marketing hype in our data world, and it’s the only day we allow ChatGPT-4 to write something on our website. But—to be honest—in terms of marketing hype language, we would never be better.