True, but that's talking about genetic algorithms, not genetic programming. I'm oversimplifying, but GA involves evolving parameters in a system to find a solution, where GP involves evolving the code itself.

https://www.quora.com/Whats-the-difference-between-Genetic-A...

Yes, I didn't quite grasp GP initially, but it simply appears to be using a GA to find an optimal syntax tree (code) that has the desired properties. From the GA perspective, wither the cost/fitness test involves executing code or not seems like just a detail of the GA optimization step.

In terms of how they behave, yeah, they're pretty similar—have a population of individuals, evaluate them for fitness, perform copy / mutation / crossover for highly fit individuals, repeat until fitness is high enough, or you get tired of it.

But, the arbitrary nature of the code being executed in GP (vs a fixed set of parameters in a GA), gives the GP system a lot more flexibility in finding an effective solution. The potential solution space is a lot larger.