"Those who forget history are doomed to repeat it." This has been said a lot recently, but unfortunately rather ironically. Here are some overarching lessons of history regarding surveillance:
1. No long-standing liberal republic has ever fallen into totalitarianism because of internal surveillance. None, ever.
2. Plenty of liberal republics have fallen into totalitarianism due to being surveilled and invaded or imperially puppeteered by foreign powers who are not and never were liberal republics.
3. The lack of preexisting pervasive surveillance has not prevented the violent fall of long-standing liberal republics.
4. The presence of pervasive surveillance by Allied governments in WW2 and the Cold War did not result in any general moral, political, or social resemblance to the other side.
5. Points 1-4 mean that the lesson of history is that surveillance is neither a handmaiden of totalitarianism nor a reliable indicator of corruption or weakness in civil society.
6. No significant surveillance apparatus was needed for the Southern US to hold millions of people in slavery for centuries and then start a murderous, devastating war killing thousands of their own countrymen in order to keep doing it.
7. Google employees know more about you than any government institution ever has or ever will, and they don't seek permission from FISA Courts or Congressional select committees before deciding what to do with the information, however flimsy that level of oversight is in the public sector. They know what your politics are, who you associate with, what you shop online for, what your hobbies and consumer preferences are, your social media content, your professional resume, your restaurant and product reviews, and what genre of nekkid pictures gives you your jollies, among many other things both more specific and more abstruse. They know this stuff in much greater detail than the skimmed averages and large-scale patterns being fed into intelligence computers. And yet despite this knowledge they are and remain just some unusually rich nerds, not gangsters ruling over their own literal empire.
8. The above points illustrate the fact that (a) knowledge is not power (which is why university professors are not Khan Noonien Singh-like superhuman rulers of the world), but rather simply one dimension in the effective application of power. And (b), surveillance does not determine or necessarily reflect the moral character and intentions of the people and institutions engaging in it.
The above are simply facts, and you don't get to have your own set of them, so if you're angry at hearing them you probably need to reexamine your own motives. However...
9. Governments that are already brutal and oppressive are more effective at controlling people if they have pervasive surveillance. So this is a real problem. But on the other hand...
10. If they're already brutal and oppressive, you're not going to be able to stop or in any way restrain them from engaging in that surveillance. The only governments you can restrain are ones over which you already have significant control. Ergo...
11. The very act of pursuing reform rather than violent revolt carries an implicit admission that the government being criticized is capable of it. So...
12. Use the fact that this is a free country to the fullest in finally deigning to address civil liberties issues that have been neglected for decades. The most urgent and important place to begin is...
13. Enforcement. I.e., what governments do with what they know, not trying to create an air-tight barrier against the flow of information that is already omnipresent due to information technology. An oppressive government that knows little (e.g., a gangster state) is just more hamhanded in its tactics; a liberal government that knows everything (e.g., Britain) is just hyper-bureaucratic. Restrain the power and restraining the rest is much less urgent: Authoritarians understand and act on these priorities in the other direction, and the only way to defeat them is if you do too on behalf of liberal values.
The Tea Party / Fascist groups would love to have a government with little bureaucracy or knowledge but a lot of itchy trigger fingers, so a lot of those people have no problem being against government surveillance. They don't need to know or care what you're doing - they are happy to let you live privately in your private poverty enforced by your Boss/Master's private business decisions, and then bury you in a private grave if your public actions are too irritating. History makes clear that brutality can and often does supply the want of information or competent governance.
So in shaping our priorities, it's important to acknowledge that what an NSA metadata-analysis computer program might know about you is slightly lower on the list of civil liberties concerns than the fact that police can stop minorities with impunity because they're minorities, shoot unarmed suspects without repercussions, lie in court without consequences, and gamble on whether judges will allow illegally-obtained information to be used at trial without the officers and agents responsible being disciplined or prosecuted if it's found to be illegitimate. Here are a few suggestions of things to accomplish that are much more central to the issue:
A. Pass laws criminalizing deliberate violation of rights in police conduct. Right now the main motivator for police to respect procedure is to avoid having evidence thrown out at trial, but sometimes they're rewarded for gambling on the indulgence of judges because the main consequence of losing is just maybe taking a hit to their career rather than facing discipline. There would be more respect for these rights if the enforcers were subject to the same criminal justice processes for breaking the law from the enforcement side as the citizenry are for breaking it in everyday life.
B. Make sure law enforcement procedures are what they should be on every level - local, state, and federal. In other words, participate in ACLU and other activist efforts to identify and close loopholes that are used to violate people's rights. This should include a requirement that information from intelligence-gathering activities that don't meet law enforcement muster cannot be used in evidence at trial, ever.
However, it would be self-defeating to erect totally absolute barriers between intelligence information and law enforcement: Demanding the government pretend it doesn't know something that it does know would be exactly the kind of surreal insanity that sours people on civil liberties and makes them think they have to sacrifice them for security. It's never very likely that a situation would arise where man-made disasters occur because of barriers between law enforcement and intelligence, and I don't for a moment believe they ever have, but every single time there is any remote excuse for institutions to claim they were hindered in protecting the country by civil liberties rules, they will make that claim no matter how far from reality it is. It's important that the policies we craft to protect civil liberties are politically strong against such lies rather than blindly promoting them.
So, perhaps intelligence information that doesn't meet law enforcement muster can be allowed to justify search warrants if the matter meets a high standard of urgency and relevance to national intelligence - but if and only if the information is declassified so that it can be subjected to the full scrutiny of the civilian justice system; was not at any point instigated or assisted by law enforcement institutions to deliberately help them get someone; and is not an egregious intrusion of national security institutions into affairs that are properly the exclusive domain of domestic law enforcement (i.e., they can't just decide to sift metadata for evidence of ordinary crimes and start promiscuously feeding that to law enforcement).
C. Close the Homeland Security fusion centers. There is no evidence whatsoever that they are of value to national security or law enforcement, and vast evidence of rampant criminal abuses of Constitutional rights.
It would be hard to argue that this stuff is not a higher priority than dealing with programs like the NSA's metadata analysis. If we achieve the above and there's still Prism, the abuses are strongly reduced while the capacity for abuse merely continues in theory, but if we ended Prism while the abuses addressed by the above continue, not much is accomplished by comparison.
Moreover, the political optics of the shrill reaction to the NSA program seems tailor-made for a libertarian rather than liberal agenda, as discussed earlier, and is thus strategically inferior as a matter of emphasis than the three points of action I identify.
PS, if you disagree with my strategic assessment, please actually make some kind of argument to that effect - don't just accuse me of being an NSA police state apologist. If you want to play that game, I can play it too: I can say you're an apologist for police misconduct, racial profiling, and DHS fusion centers if you're suspicious of wanting to focus on them rather than the NSA. But I'd rather discuss things than play the game of dueling "counterrevolutionary" accusations.