undefined | Better HN

0 pointsams61107y ago0 comments

I don't think it's made up but I think there's a good chance we're wrong about the cause. The theory is not falsifiable. There are too many interacting variables. The models have all been fit to the data and not vice versa. We've been wrong about other things that were universally accepted as true (e.g. dietary fat is bad) and later turned out to be wrong.

Doesn't really matter what I think though. I'm one person, not in a position of influence, and my beliefs and behavior on the matter are inconsequential.

I don't think that moving away from coal and petroleum is bad though. Those are dirty, polluting energy sources. I think we should be developing nuclear and electrically powred transportation and battery technology. So either way, I support moving away from burning fossil fuels.

0 comments

mirimir7y ago

> The models have all been fit to the data and not vice versa.

How else could it work? This is a system that's vastly too complicated to actually model from fundamental principles. Or at least, in any simple way. As I understand it, the components are all based on fundamental principles. But they're tweaked as needed to fit historical data.

thaumasiotes7y ago

The valid way for it to work would be to test the model's ability to predict the future. You make your models, collect data, and then reject models that couldn't accommodate reality.

This isn't done, particularly in climate science, because it's slow and difficult, but "model validation is slow and difficult" doesn't add validity to invalid models.

mirimir7y ago

Of course it's done! For one thing, you don't need to fit your model to the full historical record. So you can use more recent data to validate models fit to older data. It's like training AI.

But yes, models fit to the full historical record can only be validated in real time.

1 more reply

Scaevolus7y ago

Impugning the broad base of climate science by comparing it to some garbage results from dietary research is unfair.

windexh8er7y ago

Just to be clear, dietary fat is not good as implied by your statement:

https://www.ahajournals.org/doi/abs/10.1161/CIR.000000000000...

amanaplanacanal7y ago

Nah, that's BS. It's an argument from authority, there is no actual science done in that paper.

Now if they actually took people, and replaced saturated fat in their diet with polyunsaturated fat, holding other variables constant, then measured whether they lived longer, they'd have something. But they haven't done that.

Their reasoning is that this measure of blood lipids is associated with CVD, so if we do something that lowers that measure of blood lipids, it will also lower the incidence of CVD. That's just bad science, and they should know better.

the84727y ago

> The theory is not falsifiable.

That is fundamentally wrong. At the worst you would have to run an experiment with many earth-replicas and individual changed variables. Quite impractical but not relevant for determining falsifiability of a theory.

But more importantly models don't just have to fit data of the past, they also make predictions of the future which will automatically be tested as time advances. And pretty much all climate models do not make a singular prediction they map some input domain to an output domain. Decrease factor A by x% and the global averages of effect M will stay within some confidence interval with the mean y% lower etc.

Of course it is complex, but complexity is no excuse to throw up your arms and claim we can't possibly know anything when it's inconvenient.

j / k navigate · click thread line to collapse

0 pointsams61107y ago0 comments

Doesn't really matter what I think though. I'm one person, not in a position of influence, and my beliefs and behavior on the matter are inconsequential.

0 comments

mirimir7y ago

> The models have all been fit to the data and not vice versa.

thaumasiotes7y ago

The valid way for it to work would be to test the model's ability to predict the future. You make your models, collect data, and then reject models that couldn't accommodate reality.

This isn't done, particularly in climate science, because it's slow and difficult, but "model validation is slow and difficult" doesn't add validity to invalid models.

mirimir7y ago

Of course it's done! For one thing, you don't need to fit your model to the full historical record. So you can use more recent data to validate models fit to older data. It's like training AI.

But yes, models fit to the full historical record can only be validated in real time.

1 more reply

Scaevolus7y ago

Impugning the broad base of climate science by comparing it to some garbage results from dietary research is unfair.

windexh8er7y ago

Just to be clear, dietary fat is not good as implied by your statement:

https://www.ahajournals.org/doi/abs/10.1161/CIR.000000000000...

amanaplanacanal7y ago

Nah, that's BS. It's an argument from authority, there is no actual science done in that paper.

the84727y ago

> The theory is not falsifiable.

Of course it is complex, but complexity is no excuse to throw up your arms and claim we can't possibly know anything when it's inconvenient.

j / k navigate · click thread line to collapse