A lot more has to go right for AGI to lead to a utopia than has to go wrong for it to lead to a dystopia. Nobody knows for sure what will happen, that's the nature of the singularity, but it's easier to imagine things going wrong.

Not to mention, it's important for bad scenarios to be discussed in order for action to be taken to mitigate them. Not that many in this sub are actually involved in AI research themselves, but in general, the more people that talk about what could go wrong, the less likely things will go wrong because more people will try to prevent those things from happening than otherwise would.