I think it is just a matter of where you put resources. I am sure if you put resources into improving recycling ships some advancements will be done (it won’t be done using neural network probably).
But that’s true of everything. This guy is explicitly angry about AI not being used in ship decommission, which is just weird.
It’s not about the ships.
Shipbreaking is the author’s example, but it’s not the author’s point.
He could have bemoaned the lack of tree-trimming robots or the vaporware nature of self-driving cars instead.
The key point is the heavy investment in automating away things that bring us joy while doing nothing about vast classes of unpleasant drudgery.
Hell, look at roofers. A lot of injuries there are from falls, easily preventable with fall harnesses. It doesn’t even require a big research investment! Our society simply doesn’t value those lives enough to protect them.