I agree languages are the culprit. Too often we just develop in a new language for fun or just to fill a CV with new buzzwords. Sometimes it leads to innovation. In most cases it's pure bloat.
Recently overheard a conversation going this way:
A: "I've built a client for X in Scala."
B: "We already have one in Java. What features did you miss that made you decide to write it again?"
A: "It was in Java. Now I'm trying to write one in Haskell. It will be fun."
Even worse, that oftentimes the only way to attract new talent is by allowing folks to write stuff in a language that other companies do not (yet) have adopted. This leads to a vicious circle where new languages are going to production and the developers stay on the same level of knowledge for those languages since they do not spend enough time to dive deep enough into them as a "new kid in town" arrived.
Recently overheard a conversation going this way:
A: "I've built a client for X in Scala." B: "We already have one in Java. What features did you miss that made you decide to write it again?" A: "It was in Java. Now I'm trying to write one in Haskell. It will be fun."
Even worse, that oftentimes the only way to attract new talent is by allowing folks to write stuff in a language that other companies do not (yet) have adopted. This leads to a vicious circle where new languages are going to production and the developers stay on the same level of knowledge for those languages since they do not spend enough time to dive deep enough into them as a "new kid in town" arrived.