It's an old gedankenexperiment, the cat-in-the-box analogy, but one that has cultural tenacity for its ease of use (and, happily, a meme-friendly protagonist): if you're not looking, that cat could be alive, or not. Quantum mechanics tells us that, on a particular level, it's actually both: the position is indeterminate because it's simultaneous (in parlance, the particle inhabits a "superposition"). That is, until an observer arrives on the scene, which changes everything--we're back to that old measurement problem, one of the more fascinating ontological aspects of quantum physics, I think. (Niels Bohr had much to say about this, and I generally agree with him; also check out Karen Barad's intriguing "agential realist" interpretation.)
The salient question is this: if nothing has inherent determinate properties until an interaction occurs, what is an object prior to observation? The strange answer is a set of probabilities: Schrödinger's cat is, statistically, both alive and dead before we lay eyes on him. Similarly, a particle occupies multiple possible positions and/or velocities prior to a cut enacted by a STM or other apparatus. For ages, it was accepted as impossible to observe a particle's position and velocity simultaneously, and huge advances in q.m. have been made around this limitation. But that's all about the change: the winners of the 2012 Nobel in Physics, Serge Haroche and David Wineland--who have been friends for 25 years--have, using photons, atoms, and mirrors, discovered a way to directly observe individual quantum particles without destroying them.
The implications of this experimental coup are vast, but the most immediate application could be a pragmatically transformative one: quantum computing. Still in its infancy (Seth Lloyd at MIT has built a prototype that looks sort of like an IBM circa 1975), quantum computing manipulates quantum bits, or "qubits," at a potential rate that far surpasses classical binaries (which are restricted to being in one position at a time) precisely by tapping into the fact that, quantum mechanically, the qubits explore every possible solution to a query at the same time before reaching the correct result. The work of Haroche and Wineland allows physicists to not only observe individual quantum particles, but to control them. It remains to be seen how profoundly this affects our understanding of quantum mechanics, as opposed to its applications, but it's yet another example of how concepts previously quarantined in the metaphysical realm--like Schrödinger's thought experiment--have been successfully applied in the lab. Reuters has the money quote: "Wineland is a dedicated experimentalist, not bothered by the bizarre philosophical implications of quantum mechanics, such as the notion that reality does not exist until an observer measures it." Italics mine: the "bizarre implications" of q.m., if not relegated any longer to Schrödinger's box, are, well, catnip to philosophers of science, and underlie what could be a complete decentering of the human experience. (OK, one more comic. Originally from Analog Science Fiction and Fact.)
Scientific American has a podcast up, and some great links to early review articles etc. by Wineland and others. Also: it's textbook-y, but Haroche's Exploring the Quantum: Atoms, Cavities, and Photons is all about the gedankenexperiment made real.
UPDATE: there's some chatter about why the Higgs discovery didn't prompt a Nobel this year, but it should be no surprise: there are still data sets to analyze and findings to corroborate etc. Most speculate that a 2013 Nobel is assured for Higgs et al. (What will get interesting is which three physicists are selected for the award.) Matt Strassler says more here.