His chart is based on a single data set (the vostok ice core) for the majority of that period (plus modern temp records slapped on at the end, but it's not visible in the 10,000 year graph) which is not representative of global climate. Correlating proxy records with modern instrument records of temperature is not trivial either (which becomes far more apparent once you look at more than a single proxy record).

It's just cherry picked data to make a cheap political point. The scale of the vertical axis makes it very hard to see what is happening during interglacials, which is what we are in at the moment. If we were worried about an ice age arriving in the next couple of centuries the choice of scale might have some merit, to show how much colder it could possibly get. But we are worried about rapid heating, which that scale does a good job of hiding.