This was discussed in a similar thread a few weeks ago but for anyone who missed it the graph in the OP has little if any significance when looking at current climate trends. Past correlations of CO2 and temperature have NO relevance to current climate predictions. Climate is controlled by many factors such as the position of the continents, height and distribution of mountain ranges, planetary terrain as well as atmospheric gases. The Earth was vastly different in the past and so the CO2 and temperature correlations in the OP graph are simply irrelevant.
What is much more relevant is the climate records over the past ~1Ma years as the that is a long term trend with the planet in a comparable state to today (continents, mountain ranges ect). If you look at the data below (from the Vostok Ice Core) you can clearly see there is a correlation between CO2 and temperature over the past 500k years. Now the issue is which drives the other. The cycles seen also correlate very strongly with orbital variations known as Milankovitch cycles which alter the amount of heat the Earth receives. This causes the Earth to warm, raising the temperature of the oceans causing them to release CO2. Therefore it is likely that temperature initially drives CO2. However, the major question is whether this increase in CO2 causes a feedback effect further raising temperature thus creating a cycle of increasing CO2 and temperature. If that is the case than man made emissions of CO2 (and other gases) will cause temperatures to go up (to what degree no one can accurately predict).
