I think you mean the general class of algorithms that scale with compute times, RL being the chief example. But yes I agree to that point.