Change the display to CO2 concentration. Change the CO2 retained in atmosphere until the model (blue) matches the data (black). You have now calibrated the model to predict the concentration of CO2 in the air.
Change the display to global mean temperature. Adjust the climate sensitivity until the predicted temperatures (blue) match the observed temperatures (black). Now change the interannual variability to reflect the fluctuations in temperature from year to year. You have now calibrated the model to predict temperature changes.
Now play with the peak CO2 production year to see when we have to reverse the current increase in CO2 production, in order to satisfy the Paris Climate Agreement and keep temperature rises under 2 Centigrade. You can see what the peak CO2 production year actually does by switching back to the anthropogenic CO2 display and watching what happens as you change this parameter.
Now watch the tutorial and read the FAQs to more fully understand what you have been doing.