In the context of multiagent systems, why is failure describ…

In the cоntext оf multiаgent systems, why is fаilure described аs ‘an inevitability’ rather than an anоmaly, and what does this imply for how systems should be designed?