Even if it was true, a decade is literally a single data point in climate change analysis. Climate change is not local weather, it's not monthly predictions or even year to year values. About the smallest relative measure of time used in studying global climate change is roughly a decade. Any average data point less than a decade has a higher probability of being noise than actual average climate data. The climate models they are looking at do make predictions on the year to year stuff, but mainly as trends, but at an accuracy levels that's about as good as your local weatherman predicting weather 10 days from now. They can do it, but if you rely on it you are a fool because noise and random events have more local bearing than the trend. But once you get out to looking at climate change at the decade level the noise in the system is mitigated and the real data and trends become apparent. At that decade level the planet has been warming consistently and at an increasing rate since the industrial revolution.
And even at that the models are an estimation. There are IMO massive aspects of climate change that the models don't address because the systems aren't fully understood and in some case aren't understood at all. Inter-ocean currents that help regulate global temperatures are not understood very well, certainly not the same level as say wind patterns. Though we understand the basic mechanism we don't really understand the extent or how the climate change will affect them. As a result there are portions of the models inputs that are simply guesswork and will be refined as time goes on and more data is developed that will allow them to better tune the models. That in fact is the scariest thing about climate change, which is that our models could be completely wrong, in the wrong direction. Best case scenario is the oceans are able to absorb much of the additional heat with very little impact to global climate. Worst case is the model vastly underestimate the impacts of these inputs and in fact the consequences of global climate change are far more severe than predicted. For example, not a single model predicts much more than a gradual but small decline in the glaciers in Antarctica which will cause relatively minor sea level increases. If the models are wrong and in reality those glaciers melt, much of the worlds population is going to be displaced as sea levels increase 100's of feet (30+ meters).
Some say the models are alarmist. Others fear they aren't alarmist enough. Only time will really show how good the models are. But don't think even a decade of data contradictory to the models (not that there is mind you, that's a common myth others have addressed) is relevant, because a single data point isn't a prediction of a trend or even useful as an evaluation of the predictions. By the time 2030 rolls around we'll have tuned the models to be pretty good predictors, likely even of year to year trends. But if the models predictions are dire at that point and we haven't made reductions in C02 levels by that time, we may have already damned ourselves and our grandchildren to the worst climate change can offer. And that worst is a pretty scary future where humanity destroys itself in a fight over dwindling food and resources and displaced people.