I'm less comfortable with future predictions than with the past because there are so many variables at play. Most past predictions have overestimated the effects largely due to the earth's seeming ability to self-regulate to some extent. For instance, when ocean temperatures rise more water evaporates and creates clouds and clouds do a pretty good job reflecting the sun's rays, cooling off. Likewise as glaciers melt they often leave behind lush land that increases carbon sequestration. And if the polar ice caps melt, the earth's rotation will probably change in ways that affect how much of the sun's energy the earth absorbs.
That said, my best guess is that even if we immediately stopped emitting carbon temperatures will still rise over the next 50-100 years. I imagine it like being on the interstate and taking your foot off the gas -- you coast for a while before coming to a stop.
I think the most rational course of action is to prepare for less hospitable weather in the future. I do not see all of mankind having the will to make the kinds of changes that would otherwise be considered best practice. Part of that is basic fairness: the US and western world benefited from 150 years of cheap hydrocarbons -- we built our societies on it. Is it really fair to cut off India and China's rapid industrial growth before they get a chance to catch up?
The largest source of uncertainty is in historic temperature measurement. Mankind has had accurate thermometers for only 300 years and they were widely used only in the last 150 or so years. So scientists have to resort to proxies for temperature: tree rings, coral reef growth, ice cores, etc. These are imperfect: sometimes trees grow quickly even when it is cold (or slowly when it is hot). Each proxy has a range of error. By using multiple proxies, you reduce the errors, but the possible range of temperature is still wide.
Here is the famous "hockey stick" graph from the late 1990s with the grey/light blue zone being the extremes possible and the dark blue & green lines being predicted actual temperatures:
Attachment 47691
You can see that it is technically possible that from ~1125 to ~1175 and from ~1450 to ~1500 the earth may have had as sharp of a temperature change as it has in the last 50 years. It is unlikely - that would mean going from one extreme outlier immediately to the other - but it is within the range of possibilities.