scaling laws as comedy

[Note : imagine B is from India.]

A: Oh my God!
B: What is it?
A: I just proved this great result!
B: Really???
A: Yeah, it’s an lower bound on the achievable rate!
B: So what is it?
A: Well my scheme shows it’s at least log! Log(N)!
B: Ok…
A: Isn’t that cool?
B: Seems a bit… low.
A: Well it’s not polynomial…
B: Hardly. Log(log(N))? You’ve got to be joking.
A: C’mon! Look, if you have a log, log(N) growth you can bootstrap that up to something better.
B: No you need to get rid of a log.
A: I did get rid of a log! It’s an improvement on Singh et al.
B: So it was log log log before?
A: No, log log.
B: So what’s your contribution?
A: Well it’s log log…
B: Exactly! Log log!
A: By log, do you…
B: Log log by log?
A: No, you have an extra two logs in there, it’s…
B: 1 by log? What the heck are you trying to prove!
A: It’s log! Log! Log!
B: I give up. Why don’t you come back when you’ve figured it it out. See if you can get it to log(N). [exits]