Now I'll highlight a concern I
do actually have about a fully robotized economy: weapons proliferation. The key pieces of information needed to build a serviceable anti-tank missile or a jug of nerve gas are already lying on the shelves of thousands of technical libraries. The details may be classified, but the broad outlines in the open literature are good enough for someone with patience and a technical background to fill in workable solutions. Fortunately for civilization to date, {diligent, technically educated, sufficiently well funded, homicidal} is a set of attributes that rarely co-occurs. We've seen homicidal maniacs who were patient and well-funded enough to make their own nerve gas just once so far (Aum Shinrikyo in Japan in the 1990s).
Things change if machines do all the real work and a would-be killer doesn't need any special skills to get deadlier weapons. The unhinged man who rage-kills his ex-wife and some of her coworkers with a gun today could, in the future, ask the makerbot for kilograms of RDX or tabun instead of a gun. Outbreaks of lethal violence might be rarer, since people who are materially well-off are generally less likely to murder, but the rarer killings rooted in rage or ideology could become a lot deadlier.
In a few post-scarcity science fiction settings impulses to violence are stopped with direct human nerve implants linked to a machine panopticon that can halt dangerous actions. I don't consider that a plausible or even desirable future. In my favorite space opera setting, the Culture of Iain M. Banks, killers are pre-empted by omniscient benevolent AI oversight running millions of times faster than biological intelligence. That sounds dreamy to me, but it involves a lot of made up Space Opera physics so I don't think it is plausible. Thinking about a future where AI is capable enough to manufacture anything people ask for, but not capable enough to act as benevolent gods, leads to some odd mixtures of prosperity and catastrophe.