I have a complicated relationship with algorithmic design and modeling. I have used it for several years now. I've read through and completed most of the exercises in Algorithmic Aided Design by Arturo Tedeschi, as most interested parties have. I've integrated its use into a handful of real products. I've even written my master's thesis on algorithmic design's place in industrial design. I think it's an incredibly interesting and useful tool. It can create mind-stretching geometries, transfer data into design, and help create systems of control for personalized products. And yes, using it makes you feel like a designer from the future, here to show everyone how math and computers will lead us to the light.
However, I think I'm on the other side of the excitement curve now. The geometries are beautiful, but can lack function without directed engineering input. Geometries created by datasets are only as good as the data itself. Customization is infinite but only on the rails that you have given the system. In general, the development of these designs is time consuming and their outputs can be fickle. I'm not saying that they don't have their place and that great things can't be done in the right hands. I am saying though is that for real-world product development the situations and scope in which these programs can be utilized is narrow and can be costly to implement if not treated carefully.