It isn't novelty per se but external/sensory stimulation, which novelty can be a form of. In Eysenck's theory, everyone has an optimal level of arousal; they feel overwhelmed and anxious if it is passed and bored if it isn't met. Extraverts' brains seem to be optimized to take in, process, and respond to information coming in from the environment while introverts' brains are optimized for a slower response with more time spent on "deep processing" in the prefrontal cortex. This biological difference is thought to underpin the higher-order differences seen between extraverts and introverts: sociability, activity level/pace, assertiveness, and positive affect.
Besides novelty (i.e., perception of change in the environment), other forms of stimulation can be sheer intensity (think extreme sports and rock concerts), competition, and viscerally rewarding experiences like food and sex. Ironically, many of the sociable, outgoing extraverts seem to be quite happy with the status quo and relatively incurious; but then again, openness to experience is a separate dimension of personality in many models, and it is that dimension that captures people's tendencies to engage in intellectual pursuits, experience different cultures or more of their own culture, and try new things or question their beliefs.
You forgot Chris.
They suck, their work is lower quality, they don't solve problems well.
Chris takes a lower load of projects and they have bugs....But your job as a manager is to get enough out of Chris to turn a clusterfuck into a nuclear bomb.
Much of project management seems to be for the Chrises of our industry. The Chrises can vaguely remember how to write a for loop if they use Eclipse's auto-complete; they'll introduce at least as many bugs as they fix; and, as they grope around blindly inside the codebase, eventually, with enough management oversight and testing, they can get something approximating the specifications implemented, sort of. Yes, more than anyone, they are the cause of code rot; if there was any consistent design in the application before, it'll be gone; unit-test coverage will be near nil or covering things like getters and setters. Generic types are "too hard" to understand.
Businesses, though, usually rather hire an army of mediocre developers than a handful of good ones who can deliver better, faster.
Software production is assumed to be a line function, but it is run like a staff function. -- Paul Licker