Gen. Keith Alexander, the head of the National Security Agency, was on Capitol Hill today to complain about sensationalized media coverage and to try to justify his agency's immense data-collection efforts as necessary to fend off terrorism. He also promised that if and when his agency does something wrong, it holds itself accountable.

You know what? I don't find that reassuring. Let me give you specific reasons why:

1.) In 2002, in the wake of the Sept. 11 attacks, Admiral John Poindexter of Iran-Contra fame was appointed to create a program called Total Information Awareness, in which the new agency would collect and store any communication that was being sent by anybody to anybody else, in case that surveillance data might come in handy in detecting terror plots. The agency motto was to be "scientia est potentia" -- knowledge is power.

However, even in the immediate shock and aftermath of 9/11, the American public decided it did not want its government to have that much scientia or potentia. In the public and congressional backlash that followed, Poindexter was forced to stop down and the government announced it had dropped that initiative.

It did not. The government instead put together a nearly identical program, in secret, as part of the so-called "black budget." The full extent of that secret program did not become known until Edward Snowden did us all a favor by letting us in on what our leaders were doing in our name, and the parallels with what Poindexter proposed under TIA are almost exact.

Some national security experts say that the data-collection and data-mining system, now named PRISM, is necessary to ensure our safety. It is possible that they are right. But you know what? That is not and was not their decision to make. We had an open, public debate on that question at a time when security fears were running very high -- PATRIOT Act, anybody? -- and we decided that we preferred to take the risks rather than give government that authority.

However, We the People were overruled by They Who Know Better.

2.) In the mid-1990s, the NSA insisted on the installation of a so-called "Clipper Chip" in all electronic communications technologies so that it could intercept and decrypt messages, regardless of the code used to encrypt them. We had a pretty rambunctious public debate about the proposal, and the consensus was no, the Clipper Chip would be a step too far. But as Snowden has revealed to us, the NSA went ahead and did it anyway, largely by strong-arming computer and Internet companies to secretly give them the access that the Clipper Chip would have provided.

Again, you can make an argument that the Clipper Chip or some other means of providing that capability is needed. But again, we heard that argument, and as a nation we pretty clearly decided otherwise. We were willing to take the risks in the cause of protecting our privacy and freedom.

But again, They Who Know Better went ahead and did it anyway.

Here's the larger problem: Once TWKB assume the right to overrule us and make decisions like that for our good, where do they stop? If you study the history of technology, you soon run across something called "the technological imperative." In short, it holds that if technology gives you the power to do something, sooner or later you're going to do it. Only the most powerful of taboos -- say, the taboo against cloning humans, or the taboo against use of nuclear weapons -- can suffice to overcome that imperative, and as we've seen, even that can be dicey.

In this case, we have created the technological infrastructure to monitor every email, phone call, Internet posting, instant message or other form of communication. We are told that TWKB use that capability only in very narrow circumstances and only when absolutely necessary, and with no intent of political surveillance or other uses. At the moment, I believe that's probably  true.

However, history and human nature tell us that the definition of "absolutely necessary" always tends to broaden over time, and it does so particularly quickly when the power to decide such issues is closely held and is exercised in secret. People with great power always are tempted to believe in their own inherent goodness in exercising that power. And they always end up being wrong.

3.) Earlier this month, a demented former member of the U.S. Navy shot and killed 12 people at the Navy Yard in Washington D.C. It was not a terror attack, but for the families of the dozen innocent people who died that day, I doubt that matters much. The attacker, Aaron Alexis, had a long and disturbing record of firearm misuse and mental instability, yet somehow he retained a security clearance that gave him admission to the Navy Yard facility and the legal right to buy firearms.

We have invested tens of billions of dollars in high-tech data collection, storage and analysis, supposedly to protect our personal safety from terrorists, yet a simple background check into an individual applying for a security clearance misses important, seemingly glaring warning signs of potential danger. That suggests a couple of things. It suggests that we may be misapplying our resources in fancy gizmos and hugely expensive government contractors when those resources might better be used elsewhere.

And it suggests that even if the dangers of intense "total information awareness" in the hands of government might be overblown, so too are the promises of increased security made on its behalf. "Scientia non est sapientia" -- knowledge isn't wisdom.