While I agree mostly with what he's said, one thing I do NOT agree with is the usage of models based on current knowledge. As humans we have always have the fault of believing that things will always be as we know them, that's because our brains cannot fathom something that is much different than what we've seen. This is because we think in memories of things we've already learned. The big problem with models based on this way of thinking is that they've always been proven wrong.

We believed that CO2 was causing global warming and that exponential growth of CO2 would cause doubling of the earth temps. Well it hasn't. A doubling of CO2 has barely nudged the temperature. We've since figured out that water vapor and the sun account for more warming/cooling than all other greenhouse gasses combined. This hasn't stopped many from trying to create global policies to police CO2, even though we've already proven through ice coring that CO2 actually LAGS global temps in the cycle proving that temperature drives CO2 concentration, not the other way around, yet people are pushing "models" as "proof" that the direct ice coring evidence is somehow false.

We surmised through models that the ozone hole was caused by CFCs and that stopping all CFCs would close the ozone holes. Our models indicated both the rise and fall of CFCs and the corresponding ozone hole size and we used that as "proof" to push global CFC policies. Well, 20 years later, CFCs have dropped to record low levels and yet the ozone holes still exist even though the models said they would close up. It wasn't until recently that we discovered that they were a natural phenomenon caused by ions from the sun and temperatures near the poles..

Models of birthing rates showed in the 70s that african-americans would soon outnumber other races in the USA by the 1990's. That didn't happen. Models then showed that mexican/latin races would outnumber other races by the 2000's, that didn't happen either. Now the models show that people of Muslim faith will outnumber others in 20 years. I doubt that will happen too.

In another topic we discussed oil reserves and consumption. The models predicted the USA would run out of domestic oil reserves many times over the decades and again by 2050-ish, but our technology has evolved and we've unlocked new reserves and now the numbers show that it'll be the largest reserve in the world soon. So what happened to the models?

My point is that models have historically been used to push agenda and policy for countries and/or the whole world yet they've also historically been either proven to be totally wrong, falsified for agenda or technology has advanced to change the course of the results. I think as time goes on, we'll develop new technology that relies less on rare metals and oils and nullify these models as well. 20-50 years is a long time in technology time.

Tell Nate Silver that models don't work.

@vitaliy_kiselev I guess we'll need to agree to disagree. The DuPont money myth has been around as long as the CFC ban has and it's never had enough evidence to even be taken seriously. I've read the myths and they just don't add up. DuPont also makes the most CFC REPLACEMENT too, so it's not like they lost any money doing this. In fact, they've sold more R134 than R-12/22 simply because people felt the need to replace it. The problem with the theory overall is that nobody discovered the ozone hole until satellites imaged it in the early 80's. In the grand scheme of things, 30 years of observation is barely enough time to see a seasonal trend, let alone say that we understand anything. It's pretty reasonable to say that while CFCs drop and the ozone hole continues to drastically fluctuate with sun and temperature, that this natural cycle has gone on for millennium and we've only just discovered it. Again, people think they know everything just because that's how our brains are wired. We think we understand something and we want to change it right now, but it turns out we don't really understand it at all because we were too quick to draw a conclusion.

@brianluce Nate silver is a predictor, not a modeller. It's hard to see the differences between the two but a predictor watches human nature and predicts what will happen. Humans want to believe that we are complex and unpredictable because it makes us feel better to think we are somehow individuals, but the truth is much more boring. We are completely predictable and understandable and we are mostly alike. We instinctively know how other humans will act and we have millions of years of evolution to instill this understanding. Modelling, however, is much less intuitive since it takes theories that we have not fully grasped and tries to predict an outcome with only partial input. Add in that humans are basically scared animals and we try to predict the future in order to make ourselves feel less scared. That's the big problem with models, we tend to put too much emotion into them and even the best scientist tends to bias the data unconsciously. So you put two emotions into a model and leave out some of the data and you'll get a recipe for disaster.

This whole words have no meaning

Yep. +1

@svart

I find your post just as fucking boring way to say - motherfuckers you are wrong and I am right :-)

Some stuff about DuPont and CFC, very modest one (you can find more harsh things by yourself including some calculations of damage this hoax made)

https://eng.ucmerced.edu/people/awesterling/SPR2014.ESS141/Assignments/DuPont