In up-to-date years deep learning has accrued diverse state-of-the-art results in many fields. However, most of the research efforts chiefly focus on pure technologies or engineering optimizations of deep learning. Inadequate consideration has been paid to it from the panorama of the philosophy of technology. Given the universal Darwinism panorama, technology is regarded as the seventh life form, which, like the other six well-known life forms, has the driving forces of survival and evolution. Similar to genes, temes are the inheritors of technology. If temes do not yield to the law of "survival of the fittest", they will be discarded by the evolution. In this paper, we refined the essence hidden behind deep learning's uproar emergence from a technical panorama. We then rendered its importance and philosophical perspective of deep learning. Based on the point of view of " teme " (the basis of technology), we moreover analyzed the inherent teme -related defects of deep learning. Finally, we addressed the possible teme-driven directions for deep learning.