I have had harsh things to say about Bill McKibben in the past, but his
sit-down with Anthony Watts was apparently quite pleasant. Credit where credit is due, then. Affability isn't everything, of course, but an ability to listen and exchange is laudable.
7 comments:
Very interesting, especially the comments about what a hopeful sign it is that McKibben agrees it would be very good news if the low estimates for CO2 feedback are correct. Also Watts' (unsnarky) diagnosis of McKibben as a man who interprets scientific disputes primarily from an emotional or empathetic point of view.
I was also amused by the comments about recent "extreme" Texas weather. Having lived here my whole life, mostly in Houston, I can tell you that the intense rainfall and floods in 1978 and 2000 were much worse than the May 2015 event. Every couple of decades, some or all of Houston gets rainfall totals in a single day that can be anywhere from 10 to 35 inches. It seems unbelievable until you watch it happen, but all it takes is a tropical depression pattern.
I am a bit sympathetic to those who are intimidated by mathematical models. I used to have a child like faith in them myself, until I had to work in an environment where they were routinely used. A model is ....a tool, and an important one, in our case, used to design and engineer complex switches, switches that have multiple sensors, subassemblies, feedback loops, and whose subassemblies, in turn, have their own sensors, multiple subassemblies, and multiple feedback loops, etc. I can’t tell you how many times an ingenious, state of the art solution to a problem worked on paper and failed in the assembled product. It isn’t easy to accommodate for different materials, temperatures, pressures, energy inputs, speed, chemical environments, humidity all interacting together ………..EVEN IN A CONTROLLED ENVIRONMENT, where the variables are known. I have observed how a seemingly minute change in value can be amplified as it works its way through the system, with unintended consequences (we called it a cascading effect) and vice versa.
The difference between us and academics is that our models eventually had to consistently match, the customer’s required spec, in the real world, or we didn’t get paid. It takes real hubris to insist you know what you can’t know. In any case, it can’t be said often or emphatically enough, a model is a TOOL, not an end in itself.
Continuation of above post:
On the other hand, if we could have cherry picked the data, we would have succeeded every time. But alas, in our world the model had to fit the data, not vice versa.
Every time I run into someone who expresses this basic thought, that models have to be fixed when they don't match data (not vice versa), I breathe a little sigh of relief. In the last decade or so, there has been a weirdly disorienting move away from this concept among people of good education who really should know better. What must be going on in some parts of the scientific research world? I don't even think it's purely a question of partisanship, or at least not conscious partisanship. Is it some kind of conviction that one's science has to be socially responsible? It does, of course, but it's sad that such an idea would be played out in this way. It's one thing if you're being to forced to develop weapons for Nazis at gunpoint and you fudge your data on purpose, as a kind of sabotage . . . .
People so seldom change their minds. And seldomer yet in the course of an argument.
Take every chance you get in life, because some things only happen once
Obat Karurawit / Karurawit Herbal, 9 Jenis Bahan Makanan Alami Untuk Mengobati Flek Paru-Paru,
trung tâm tư vấn du học canada vnsava
công ty tư vấn du học canada vnsava
trung tâm tư vấn du học canada vnsava uy tín
công ty tư vấn du học canada vnsava uy tín
trung tâm tư vấn du học canada vnsava tại tphcm
công ty tư vấn du học canada vnsava tại tphcm
điều kiện du học canada vnsava
chi phí du học canada vnsava
#vnsava
@vnsava
Post a Comment