Compared to GHC, MRI really is primitive. Parts of this are intentional: the object model is intentionally simple, and one of the design goals was to keep unixy goodness close to the surface, so the IO model, forking and so on are quite shallow shim layers over the system calls.
Other parts, like the threading, the GC, and the parser (oh god, the parser) are suffering from baggage that's never had GHC-levels of academic effort thrown at it.
To the contrary, Ruby, probably more than any other language, has a large community
that's highly enthusiastic about its programming style and culture, which is the reason
most of those libraries and ecosystems were created in the first place. I know this is just
a few bullet points, but vaguely mentioning a few (good) reasons to evangelize Haskell
(plus a motorcycle) along with three seemingly random relatively small issues as
criticisms of Ruby (plus a horse and wagon) distorts reality almost to the point of
unrecognizability
The presentation is making the point that MRI itself is primitive, and addresses the plus points of the community in the very next slide. For people who care about the implementation underlying their code as much as they care about their own code (or more), MRI's shortcomings are a visible downside.
Other parts, like the threading, the GC, and the parser (oh god, the parser) are suffering from baggage that's never had GHC-levels of academic effort thrown at it.
The presentation is making the point that MRI itself is primitive, and addresses the plus points of the community in the very next slide. For people who care about the implementation underlying their code as much as they care about their own code (or more), MRI's shortcomings are a visible downside.