Why are table$-$driven agents often impractical for complex environments?

Select one:

They cannot handle stochastic environments.

Their table size grows exponentially with the increase in percepts.

They are too slow in decision$-$making.

They require real$-$time learning.