You don't need a 'conscious' observer to force the system into an eigenstate. An inanimate device recorder, or piece of photographic film, will do the same.
Quantum systems obey an entirely determinate set of laws. All sorts of funky things happen when you try to enforce 'classical reality' onto quantum systems; in the particular system I'm working on right now, a hydrogen atom occupies two sites on a particular symmetric molecule. The ground state wavefunction (which the molecule occupies at low temperature) has equal amplitudes on both sites. The atom, if you like, is in both places at once. If you try to force it into one place, you can do that, but once you 'let go', the atom will oscillate periodically back and forth between the two sites, with considerable higher energy than that which you get if you let it alone. You can measure that frequency (we just measured the largest such yet recorded; it's off to Science this week). In fact everything about the behavior of the system is predictable; much of it just has no analog if you consider atoms to be billiard balls and potentials to be hard surfaces.
Most of the 'measurement paradoxes' seem to me things that are counterintuitive mostly to people who aren't used to thinking in quantum terms. The EPR paradox/Bell's inequality stuff can indeed get very counterintuitive, but it doesn't contradict any physical laws.
The Turing test (different from the Turing machine) says that you can define a machine as conscious if a conscious oberver cannot determine by interacting with it that is simply a machine. The famous Eliza program, for example, might be an attempt to construct a conscious being by a Turing test. I find the test distinctly unsatisfying, but can do no better.
It is because consciousness is not a well-defined scientific term at the present time that these three scientists are investigating it, hopefully to be able properly to define it. Do you think this is an illegitimate endeavor for science?
I think it's an unscientific endeavor. We don't spend much of our time defining things. One of the few matters on which I still agree with Popper is on the definitional problem. Popper said we shouldn't waste our time on definitions, as definitions are shorthand for lists. So, when you say human, you mean the list {yourself, your mom, your dad, ...Mao Zedong...}. When you say 'conscious', what is that shorthand for?
For Lurkers, here's the last bit I have found on Bell's Inequalities:
THE FIRST ENTANGLEMENT OF THREE PHOTONS has been experimentally demonstrated by researchers at the University of Innsbruck (contact Harald Weinfurter, harald.weinfurter@uibk.ac.at, 011-43-512-507-6316). Individually, an entangled particle has properties (such as momentum) that are indeterminate and undefined until the particle is measured or otherwise disturbed. Measuring one entangled particle, however, defines its properties and seems to influence the properties of its partner or partners instantaneously, even if they are light years apart. In the present experiment, sending individual photons through a special crystal sometimes converted a photon into two pairs of entangled photons. After detecting a "trigger" photon, and interfering two of the three others in a beamsplitter, it became impossible to determine which photon came from which entangled pair. As a result, the respective properties of the three remaining photons were indeterminate, which is one way of saying that they were entangled (the first such observation for three physically separated particles). The researchers deduced that this entangled state is the long-coveted GHZ state proposed by physicists Daniel Greenberger, Michael Horne, and Anton Zeilinger in the late 1980s. In addition to facilitating more advanced forms of quantum cryptography, the GHZ state will help provide a nonstatistical test of the foundations of quantum mechanics. Albert Einstein, troubled by some implications of quantum science, believed that any rational description of nature is incomplete unless it is both a local and realistic theory: "realism" refers to the idea that a particle has properties that exist even before they are measured, and "locality" means that measuring one particle cannot affect the properties of another, physically separated particle faster than the speed of light. But quantum mechanics states that realism, locality--or both--must be violated. Previous experiments have provided highly convincing evidence against local realism, but these "Bell's inequalities" tests require the measurement of many pairs of entangled photons to build up a body of statistical evidence against the idea. In contrast, studying a single set of properties in the GHZ particles (not yet reported) could verify the predictions of quantum mechanics while contradicting those of local realism. (Bouwmeester et al., Physical Review Letters, 15 Feb.)