Well, as neutral as can be for someone who put lots of effort into ML way back in 1980 for a while there before the crash.
That's a pretty long time to be considering things like this, but it's so far removed from today that by now more decades have been spent as an outside observer with a bit of a unique background than anything else.
On any one subtopic I may still have a knee-jerk reaction more positive or negative than neutral, where I might not be that well-informed or have a full grasp of the implications. Regardless of how flexible I am as things emerge faster than ever, at times it could just as easily be lumped in with either the doomers or those who feel AI can do no wrong.
Maybe a few decades is not enough to come up with the really hard answers. Or even all the right questions?
I'm first to admit that about the main thing I do is respond to prompts for more than a decade without even trying. Without any AI at all and that's not likely to change.
If that's confusing, oh well.
OTOH the huge advances in language models make it easier for non-English speakers to make their messages more readable which can help a lot. And with computer languages being languages themselves progress here is remarkable too.
(* maybe I shouldn't use that word!)