Ah, if only I had headphones or speakers on my work pc, I'd be able to listen to this. As it's Randi, I imagine it'll be interesting and funny. And, what do we really need to know about astrology, beyond the fact that it's a big pile 'o' crap.
While you are absolutely correct to point to confirmation bias, the idea that this says anything about astrology is absurd. The ability to construct a description that many people feel accurately describes them says nothing about the method you claimed to have used to generate that description. This is a typical red herring. This experiment could just as easily have required the participants to undergo several hours of psychoanalysis with a qualified professional, or had them fill out a Myers-Briggs test, and the ability to provide a seemingly correct description would have said nothing about the accuracy of those methods.

Another comment says "what do we really need to know about astrology, beyond the fact that it's a big pile 'o' crap". This statement shows that any association of this experiment with the accuracy of astrology is actually an example of the widespread confirmation bias against astrology. Because the predominant scientific view presents no feasible mechanism through which astrology could function, it is assumed to be false and is therefore not even considered worth investigating. This is a very limiting perspective to hold about any topic.
This demonstration isn't a scientific experiment purporting to disprove astrology. It is a demonstration of why so many people untrained in critical thinking are convinced that astrology works.

That astrology is fallacious is, as you correctly point out, assumed for this demonstration. But you ignore the fact that there's already a mass of scientific research showing that astrological predictions do no better than chance.

It's pretty dumb to claim that because THIS demonstration doesn't show that astrological predictions fail, science CANNOT demonstrate that astrological predictions fail.
The statement accompanying the video was "Everything you ever needed to know about astrology." It was this to which I was responding. You point out that there's more to know, in particular considerable research on the topic. I absolutely agree that any and all research should be considered when forming an opinion.
Humor
My "everything you ever needed to know" comment was meant tongue-in-cheek. I'm quite familiar with that standards of reproducibility, falsifiability and causality and I would not even begin to presume that this short, amusing clip meets any of those standards. Hence, the reference to "confirmation bias".
Re: Humor
Ah, of course. I laughed myself at the looks on their faces when they realized what had been done. I meant no offense, and I apologize for forcing you to explain the joke, that's never pleasant. :)
Nice turnaround, that.

However, if we're sticking with the scientific method here, we don't assume that coincidence indicates causality in the absence of reproducibility- in this case, unless there is a reputable, peer-reviewed, non-biased scientific study of which I have heard nothing indicating that astrology is better at predicting the future / personality characteristics / relationships / etc than it statistically should be, there's nothing to lend astrology weight.

If it was just a matter of not being able to figure out how it could function given our understanding of physics, that would be one thing- the laws of physics have been disproved and amended accordingly before, and I'm willing to bet that the will be again many times.

In short, it's not the lack of an explanation that is the killer, but the inability to reproduce reliable results that makes astrology and the scientific method incompatible.
You make a very good point, and indeed my statement about a lack of mechanism is insufficient to explain the prevailing view of astrology.

The only other question I would raise is in regards to your statement about a "reputable, peer-reviewed, non-biased scientific study". According to the standard approach to scientific publication, the peers would have to be in the same field as the research and the researcher. You wouldn't ask a chemist to review a paper about astrophysics. How many of the researchers publishing about astrology would consider themselves astrologers? How many of the reviewers were astrologers? My understanding is that the answer to both of these questions is "not many". I understand, though, that there are some, and that they too may have failed to show correlation.

I suspect that any paper published by an astrologer and peer-reviewed by others who consider themselves astrologers would immediately fail to satisfy the requirement of "reputable" for the majority of the scientific community. This just brings us back to the idea of confirmation bias.

I'm not trying to convince anyone of the validity of astrology in particular. Rather, I think there's an important distinction to be drawn between the current beliefs of the scientific community and the scientific method itself. As you said, the laws of physics have been rewritten many times. The points I am trying to make would apply to any system that falls outside of the scientific mainstream.
I'm going to jump tracks here a little bit in my reply.

My fiancee's father is a doctor- a practicing psychiatrist working mainly with a county mental health agency near to where he lives, to be precise. Graduated cum laude from Yale, got his MD at Case Western, and has been in practice for over 30 years. In short, he's a fairly reputable source from a scientific perspective, at least as regards mental and to a lesser extent physical health.

At the same time, he is, by his own description, a shaman. (He started out nearly as skeptical as I am, to be clear.) In his own private practice (not so much his work for the county), he employs methods like energetic healing, spiritual journeying, and depossession with his clients. According to him, these methods have a definite impact, regardless of whether or not he understands all of the underlying mechanisms.

To my knowledge he has not performed any studies to demonstrate conclusively that these methods have more than a placebo effect- for him, his personal experiences are enough to convince him that his methods are effective. As I don't share that body of experience, I don't share his conviction- I think that's fairly understandable.

However, he's the sort of person I would *want* to perform such a study. I'm sure there are astrologers who would fall into the same category- people with scientific training and credentials who still believe in XYZ unexplained phenomena. And once those people have performed the study and shown results, let skeptics like me at the data to try to recreate it; if a non-believer can reproduce it, then we have some hypothesizing to do.

If we can't, though... If we accept the validity of both studies and ascribe the differing results to confirmation bias, we're left at loggerheads and without even a hypothesis for how the believer could be right and this apparently irrational phenomenon could be taking place.

Little wonder that, in such situations, the skeptics usually win out...
Ah, just watched that vid at home with sound. Derren Brown did a similar trick, which he documents in his book Tricks of the Mind, which includes the actual horoscope reading he handed out.