We need to think about what AGW actually "predicts".

In detail, it depends on how big climate sensitivity is, and also what time period we are looking at.

Remember that if small enough, the AGW impact will clearly be trivial compared to natural climate variations and weather. Also, even if the overall impact is big, the delta impact between two very close dates (AGW impact at 2012 minus AGW impact at 2011) is always going to be trivial compared to natural variations.

Qualitatively, over short enough time periods, the AGW signal will always be masked by the natural climate /weather noise; whereas over longer time periods, the signal will be evident as a statistically significant trend imposed on top of the noise.

Questions:

1. Do you disagree with this qualitative prediction, or do you really think that AGW predicts that every year should be hotter than the previous one (so each year is a record high)? Do you think it even predicts that every day should be hotter than the previous one (so each day is a record high)? Clearly if your mental model of AGW is "AGW predicts year on year increasing temperatures; we don't observe that, therefore AGW is falsified" then this is a rational deduction, but from false premises.

2. Would you like to make a bet on the average global temperature in the decade 2010-2019 versus average global temperature in the decade 2000-2009? If you really believe there is no global warming trend, you should be happy to take an even money bet on this decade being cooler than the last one. Fancy the bet? (Hint, you would have lost a similar bet made in any previous decade since the 70s).

3. If your answer to question 2 is "Ahh, I think I'd bet on warming after all, but I don't think it is anthropogenic" then you're acknowledging that the argument about year to year temperatures (less than decade averages) is basically irrelevant. So why raise it? If this were a political forum, and you were trying to score points in a debate, I'd understand it.... But this is Less Wrong.

4. If your answer to question 2 is "Climate scientists/meteorologists might end up reporting that each recent decade is hotter than the previous one, but I won't believe them because I think they're lying" then again you're acknowledging that an argument over year to tear temperatures is irrelevant (fictional/made-up evidence is irrelevant to choosing between hypotheses). So again, why raise it except for point-scoring?

I say, you are lucky to be in the environment where the truth is leaving anyway. But that was a pure coincidence and you managed to neutralized any conformation bias from this.

Irrelevant. AGW is not some niche belief where evidence is hard to come by. It is a subject of massive scientific scrutiny. So I do what we should all do in these cases: accept the scientific consensus, while widening the uncertainties. Confirmation bias may have made this easier for me, but it was the right thing to do. When you can slap me down for being biased, is when I use weak excuses not to go straight to the scientific consensus. I try not to, but I'm sure I've got blind spots.

No, I don't buy the AGW story. I don't find it particularly probable, since we have no temperature rise in recent years, predicted by it.

Do the terms "hottest decade on record" ring any bells? (See for instance http://www.cru.uea.ac.uk/cru/info/warming/, but feel free to use any large scale temperature record) These scientists have been predicting rises since the 60s, and have essentially been correct.

And even if it was true, it wouldn't be a very important matter.

By the scale of the existential risks I normally deal with, I agree. On a normal human scale, and in terms of the costs it imposes, somewhat important.