Failure is a Must

This week I disproved the very premise of the work I’ve been doing for the past four months. It’s all bullshit, and now there will be no paper and likely no conference in Poland. The worst part is that in retrospect the problem is so obvious it hurts. I took the boss a one line proof that we both stared at in silence for several minutes, defeated, and then with shrug I get, “Well, there’s an ample supply of stupid. No getting around that. All we can do is try to tap into the scarce supply of smart.” Your average Ph.D. will admit that they could reproduce all of the work in their thesis in about 6 months time. The other three and a half years are devoted to figuring out exactly what work to do, and it’s a painful process of trial-and-error. I guess I’ve got four months of that done; could have been more.

Yesterday one of the campus bars closed down and there was a party to finish the kegs. I met some random crew of mechanical engineers who were over the top. These guys built a robot that mixes drinks in precise proportions and volumes, and wrote a program that uses machine learning techniques to invent new drinks based on people’s ratings of its previous efforts. They were drunks, every one of them, and they loved to sit around and elaborate on ridiculous ideas, like getting ransom from NASA on the threat of defacing historical moon landing sites. We’re to imagine a rover with a mounted feather duster poised delicately over Neil Armstrong’s footprints. For the first Friday in a long time I was entertained instead of entertaining and it was good.

Thanks to whoever found Babylon Circus. I daresay Au Marche Des Illusions is as good as Underdog World Strike.