On the issue of safeguarding nuclear reactors and long-lived nuclear waste into the future.
The nuclear enterprise entails the handling of significant quantities of hazardous materials. Trust of the establishment and of stable governance into the future is certainly an issue. How can safety be ensured through political unrest and other significant governance changes that might happen at sometime in the future? We can ask that question of many of the legacies that we are leaving the future generations - fossil fuel depletion, pollution, toxic waste (that has an infinite half-life), social inequity and injustice, billions in extreme poverty, etc. In the case of nuclear power and radiation, it is the risk of people dying.
So let's dig a bit deeper into which human deaths are more important to you? Future generations? Current family and friends? People currently living in far off places? I ask this because if the objective is to minimize the number of deaths for a given path forward chosen then we need to know how to estimate the death count for each postulated path forward, ie how to integrate risk over space and time. For starters, we must apply a discount rate to future deaths (that may or may not happen). Bird in the hand and all that. At a modest 1% / yr discount rate, discounted for 500 years, we get 0.99^500 = 0.01, implying we should spend 100 times more on today's nuclear waste risk than we should on the risk 500 years from now. Fear mongers would have us use a discount rate of 0%, which leads directly to nonsensical results. If I don't discount the future to some degree then saving ALL future generations (from hypothetical deaths) dominate over the much lower numbers at risk today. We'd end up spending all our resources in prevention of distant future hypothetical deaths rather than focusing on immediate needs. For similar reasons we care more about those closer to us physically and emotionally. Presumably people close to you are more valuable that distant people you don't know. So there is a space discount rate too. Uncomfortable with making a risk calculation that requires putting a value on human life? We do it all the time because we must. Should I buy a Prius to step lightly on the earth thus helping to save the planet for future generations or should I drive a more substantial vehicle to lower the risk to my family?
We can't be absolutely sure about future political follies so we can't be sure that what we create today will not be misused in the future. All we can do is weigh the odds and safeguard as best we can . But are you sure that picking a nuclear free path will lower the projected number of deaths? See The Risk of Energy Production. Solar, wind and hydro are not benign. Nor can they meet energy demands worldwide. So 'no nukes' means more fossil which means more deaths. Nuclear waste can be safely and securely stored and would hardly be the preferred terrorist weapon given the ready availability of conventional weapons, including biological weapons. We also have a pretty good handle on real-world death rates when nuclear goes terribly wrong. The actual body count is actually pretty low compared to the wild estimates of the fear mongers. TMI and Fukushima death totals? Zero. Chernobyl? 31. Don't believe the wild Chernobyl estimates you hear about; they are bogus numbers based on the misuse of the Linear No-threshold Theory (LNT) since there are no observable negative effects due to low doses of radiation over large populations.
I have trouble too with the logic of rejecting a path because it has some negative aspects. In truth, there are a lot of other aspects that need to be factored in before a conclusion on appropriate actions can be reached. Quantification and comparison of risks and benefits are needed to reach a conclusion. Some people have argued for the precautionary principle, which states that even if we are wrong in our fears the consequences are so large that we should act to protect ourselves against the possibility. That's fine as long as the opportunity cost is acceptable. In this case, this is neither evident nor has it been demonstrated.
It seems pretty unethical to me to reject nuclear knowing the alternatives will lead to far more deaths based on the data we have at hand (comparative death rates from various energy sources). Personally, I am more moved by actual death rates and sound projected rates than I am by imagined nightmares.
So in short, we don't need to trust in stable and predictable governance so much as we need to make proper estimates of the real risks and design in the appropriate safeguards. Let's empower people (and we need power/energy to do that) to raise the standard of living worldwide so there is a reduced need for violence, let's empower women worldwide, let's promote fair energy trade and energy independence (read: get out of the middle east), etc. But let's not dismiss nuclear because of an unquantified fear.