The
Meditations are arranged as a series of "trials" one must pass before graduating on to the next. In his
Replies, Descartes clearly states that deductive "arguments" are basically of no use for the type of knowledge he is after. The famous cogito ergo sum is not meant to be a syllogistic argument (i.e., should not be represented as kmik has done

), and is not formulated as such in
Meditations. Rather, this "truth" is intended to register "intuitively" by "the light of nature" when one is along the meditative path opened by the philosopher. Descartes says an infinite regress of justification and "proof" cannot provide
any knowledge, and that even our most rigorously worked out arguments all rest on some sort of intuitive apprehension (the elusive "grasping" or leap of understanding).
Regarding Blowtus' concern of the validity of the scope of representation (i.e., correspondence of "internal" or "mental" representations): For Descartes, the rules of thinking ("logic") were held to be universally valid. Thus, whether the objects of sensation correspond to anything "real" (i.e., mind independent) is initially doubted, but not the scope of human reasoning and logic. Both (correspondence of sensory representation and logic) are claimed to be secured by the demonstration of God (who would not deceive us in either sense).
What I am trying to stress is that "cogito ergo sum" makes little sense when extracted as an isolated argument, as its demonstration is tied to a certain method and a certain understanding of scholastic philosophy and theology.