I'm talking about the encryption. If you have a poorly configured security and/or a weak passphrase to the key, then yes, you might be able to extract it from a special sniffer and use highspeed computing to crack the key.
The goal is to put a good key in that would keep the bad guys guessing for a couple of years. There are also web-sites that you can test your password to the key, to see how long it would take to crack using the best current technology.
As far as WPA2 this site is a good example:
http://forums.macrumors.com/showthread.php?t=429524
Member parrotheadmjb had this to say:
"I hate to say it, but im pretty confident that this is fake, i tested (under the presumption it was fake) and it generated the same key as the one in the video, not only was the key the same, but EVERY LINE in terminal was the same that appeared in the video. even if it did "crack" the packet and gave you this "key" it would be useless since you have to inject packets into the router to hack a wep encryption, the ipod touch does not have the hardware, or ability to perform the required injection, rendering this useless. I still have no idea why someone would go through the trouble of writing a fake script to run in installer and post a video of it on youtube.
Even if wep was possible, wpa-psk or wpa2-psk would be IMPOSSIBLE to perform on the ipod touch, i have heard people talk about vulnerabilities of tkip in wpa-psk but wether or not you can hack it, is unknown to me, but the cpu, etc required to hack wpa-psk is far more than the touch can provide. wpa2-psk is out of the question, it implements the far more secure AES encryption, your computer cant do it - your touch definitely cant, unless you run a password "cracking" software to use dictionary attacks, brute force, etc."
This is pretty much the case everywhere using other hardware. Now if you have a pretty good passphrase, it would take some heavy duty equipment to crack that, unless your hardware wireless device has a circuit vulnerability. That is not the fault of the encryption protocol, but the wireless device.
I've tested the WPA2-PSK on my clients and it would usually take 11 years to crack the average good paasphrase used by my clients. Now as computing power goes up, the likely hood WPA becomes obsolete, also goes up. But you got to balance this with the fact that some people don't live in crowded areas and would find out the cracker in a heart beat! Plus, the majority of my clients don't have either service or personal information worth going to that much trouble to gain.