Has anybody desired using chancejs like this: Chance.string(); rather than const chance = new Chance(); chance.string();

When unit testing in a JavaScript/TypeScript project I have found great value in using chancejs (https://chancejs.com). But in each different spec file I have to instantiate a new instance of chance like so:

Example with instances:

//Spec 1
const chance = new Chance();
const personMock = {
    firstName: chance.name(),
    age: chance.integer(),
};

//Spec 2
const chance = new Chance();
const addressMock = {
    address: chance.address(),
    isApartment: chance.bool(),
};

I have not discovered a need for a chance instance, except when a known seed is declared (https://chancejs.com/usage/seed.html). In all cases where it is not desireable to use a seed, it seems like instantiating chance is only extra overhead.

Question 1: Has anybody desired using chance in the same way we use Math.random()?

Example without instances:

//Spec 1
const personMock = {
    firstName: Chance.name(),
    age: Chance.integer(),
};

//Spec 2
const addressMock = {
    address: Chance.address(),
    isApartment: Chance.bool(),
};

Question 2: Can you foresee any drawbacks to doing so?