I don't know about that.
While part of me is pretty sure that we may have trouble recognizing alien life when we eventually do find it, another part of me wouldn't be at all surprised to run into something vaguely human in appearance, thanks to developmental biology.
Yes. Convergent evolution, basically.
A limb is a limb and there will most likely be some kind of symmetry at play. At some point in the evolution of life on a planet, natural selection will favour a certain kind of symmetry for the dominant clade (radial or bilateral symmetry being the most obvious options). On Earth, it ended up being radial symmetry for almost all, with exception of things like sea stars or jellyfish, and very simple animals like polyps or single cell organisms.
I guess it would be possible for an intelligent species to evolve from a sea star -like ancestor with radial symmetry, eyes everywhere on its skin, and some number of radially symmetrical limbs - but just as likely (and possibly more so) would be a bilaterally symmetrical species with mirror-symmetric limbs.
And a bilaterally symmetric species almost certainly has a head of some kind - whether the bulk of the central nervous system (brain) is stored in the "head" or not, that's a different question. For all we know, brain could be in the place of the liver or somewhere else in the abdomen in some species (although most likely visual, auditory and olfactory sensors would be in close proximity of the brain to minimize signal latency)
The number of limbs would depend on, again, how the evolution happened to go on that whatever planet, and on bilateral animals the number of limbs would be even. And the thing is, any number higher than four is probably superfluous. Sure, on arthropods we see a lot more limbs, but more complex animals would have to spend a lot of resources to grow limbs that they don't, strictly speaking, need - four limbs should be enough for any aquatic or terrestrial species! Maybe six limbs could be a possibility, if the presence of a surplus limb pair was somehow beneficial to the early common ancestor of that clade, but I would rather put my money on four being the most common number of limbs in complex animals, in most planets of the universe. Though I have to admit, six-limbed animals would be particularly interesting because you could then have a quadruped animal with an extra pair of limbs for other task - either for manipulation, like the shifty looking cow or centaurs from our mythology - or maybe he extra limb pair would be evolved to provide flight, like the Pegasus or griffins or dragons even.
In terrestrial sentient animals, one limb pair would almost certainly adapt to being used for manipulation, and on four-limbed species, that would almost certainly mean partial or fully bipedal locomotion, in which case you already have a basic humanoid body type (joint number and directions and body proportions notwithstanding). Six- or more-limbed species could either retain quadruped locomotion and have one pair of manipulator limbs, or they might also develop bipedal locomotion but have two pairs of manipulator limbs.
To summarize - bilateral symmetry, a head storing brain and visual, aural, and olfactory sensors, four limbs, two for manipulation and two for bipedal locomotion - a humanoid form - might actually be a fairly common basic "blueprint" in sentient terrestrial species.
On the other hand, aquatic sentient life wouldn't necessarily follow these rules (see: cetaceans and cephalopods). Bipedal locomotion certainly wouldn't be particularly useful for them...