Climate change, GIGO
Note, this conversation may or may not ever have happened.....
Scientist: " I need a computer simulation"
Programmer: "Okay, no problem, what do you want it to do?"
Scientist: "Okay, I'll give you data, you use that to extrapolate how that data will progress"
Programmer:"Right, so to just to confirm, you'll give me detailed data and I then compute how that will progress over time?"
Scientist"Yep"
Later......
Programmer"Okay, i've done your algorithm, lets get the data in"
Scientist"Right the data is temperature figures and CO2 figures for the last 100 years"
Programmer"Cool, in it goes. Execute. Great, your data for the last 100 years show a steady increase, my algorithm shows more increase"
Scientist:"Oh my god. Its as i expected, we're all doomed"
Programmer:"Well we could double check, have you got more data? We're talking a long term thing here aren't we?"
Scientist"Er no not really, sorry"
Programmer"Riiiight. So we're talking about a data model that at least spans billions of years, countless variables and we only have a minuscule fraction of data from that whole and yet this model is expected to accurately predict how the model will continue over time? In that case I'd say the outcome should be it'll be warmer, or colder, or more likely both. The model can only work with the data you give it. Garbage in, garbage out."
Scientist:"Ah sod it, we'll publish anyway".
Biting the hand that feeds IT
