I Agree With the Random Objective, but It Doesn't Write Objective.

I think there's something kind of fundamentally wrong with that because it's not optimized. It doesn't write objective. I feel like I mean, I think it's going to be discounted very easily. Maybe there's better ways to do out there like that and maybe some of the reinforcement learning start techniques would like help here. Yeah. Interesting. Do you feel like you had any opinions that you used to hold strongly? But now you've reversed your position? Something that I do not hold strongly at all.