Does this graph take into account the gradual immersion of many once isolated measuring locations into urban heat islands? I wouldn't take this at face value because of the difficulty in measuring 'global' temperatures over time, especially when you consider the locations of many of our weather stations today and the absence of weather stations in many spots a century and a half ago.
What's the theory behind this metric? Is it an average taken at a particular time of day, an average of the daily highs, an average of the daily lows, an average of the sum of the daily high and daily low, or some other calculated value? Each method has serious flaws even if the data were valid.
Do you give equal weight to stations regardless of station density, or do you calculate an average areal temperature from the number of available stations irrespective of the number of stations thus ignoring the impact of sample size on validity? Again, your methodology significantly affects your results.
That said, yes, the average temperatures we measure are most likely slowly rising. However, I suspect we differ regarding probable causes and possible effective interventions. Were we to abruptly wipe mankind from the planet, would these changes slowly cease and then reverse? I don't know for sure, but I doubt it.