Emissions are growing exponentially, and are accumulative. When you look at the CO₂ footprint graph, the damage done to the climate is essentially proportional to the area under the curve. What 30 years ago was just a concern of some scientists, today is becoming visible to people, and 30 years from now will be obvious (under current growth, between today and 30 years from now, we'll have emitted more CO₂ than we had between today and the dawn of mankind - a "neat" feature of exponential growth).

In the real world, exponents don't last forever. The economy will eventually give in, the question is how. 60 years ago the bet was that we'll run out of resources to burn. Today, the consensus is that we'll most likely destabilize the climate to the point of total civilizational collapse before we exhaust economically exploitable fossil fuel deposits.

As for the next years being critical in determining how the world will look like many decades from now, keep in mind that the damage CO₂ is causing is proportional to the integral of the yearly emissions curve, and also that feedback loops take time to settle at a new equilibrium. Models predicting average temperatures and sea levels take into account that even when we become completely net carbon neutral, the already accumulated CO₂ will still warm up the climate for decades. So it very much matters when we get to carbon neutrality, and the sooner we get there the (exponentially) better.

So it's within the next 50-100 years and on the list of things we should be worrying about.