Despite Otto Warburg's 1931 Nobel Prize for his work affirming the role of metabolism in carcinogenesis, there has been little further interest in this association between metabolism and cancer. Disinterest has, in part, been attributable to the notion that Warburg's description of a relation between a shift to glycolysis in carcinogenesis may be an epiphenomenon rather than a mechanistic determinant. By studying the critical cellular energy sensor AMP-activated protein kinase (AMPK), I postulate that the association between intermediary metabolism and tumours varies over time. Through accumulation of carbohydrates and pan-inhibition of AMPK, premalignant tumours may gain a replicative advantage through the repression of senescence. Conversely, malignant tumours, with a defective tumour suppressor contingent, undergo a "glycolytic switch", in part by tolerating a degree of AMPK activation, to mitigate substrate limitation. I contend that this Janus-faced relation with intermediary metabolism contributes to carcinogenesis; if proven, this finding would have important implications for public health, in that it would lend support to the idea that prevention of obesity, and caloric restriction and exercise could reduce the predisposition to cancer.