Never ascribe to malice that which is adequately explained by incompetence.
― Robert J. Hanlon
I’ve written mostly about modeling and simulation because that’s what I do and what I know best, but its part of a larger effort and a larger problem. I work for a massive effort known as science-based stockpile stewardship where modeling and simulation is one of the major themes. This whole effort was conceived of as a way of maintaining our confidence (faith) in our nuclear weapons in the absence of actually testing them. There is absolutely no technical reason not to test them; the idea of not testing is purely political. It is a good political stance from a moral and ethical point-of-view and I have no issue with taking that stand on those grounds. From a scientific and engineering point-of-view it is an awful approach, and clearly far from optimal and prone to difficulties. These difficulties can be a very good thing if harnessed appropriately, but today such utility is not present in the execution of our Lab’s mission. As one should always remember, nuclear weapons are political things, not scientific, and politics is always in charge.
The science-based stockpile stewardship program is celebrating its twenty-year anniversary. Our political leaders are declaring it to be a massive success. They have been busy taking a victory lap and crowing about its achievements. The greatest part of this success is high performance computing. These proclamations are at odds with reality. The truth is that the past 20 years have marked the downfall of the quality and superiority of our Labs and the supremacy of these institutions scientifically. The program should have been a powerful hedge against decline, and perhaps it has been. Perhaps without stockpile stewardship the Labs would be in even worse shape than they are today. That is a truly terrifying thought. We see a broad-based decline in the quality of the scientific output of the United States, and our nuclear weapons’ Labs are no different. It appears that the best days are behind us. It need not be this way with proper leadership and direction.
Confidence is something you feel before you truly understand the situation
― Julie E Czerneda
Nonetheless given the stance of not testing we should be in the business of doing the very best job possible within these self-imposed rules (i.e., no full up testing). We are not and we are not to a relatively massive degree. This is not on purpose, but rather by a stunning lack of clarity in objectives and priorities. We have allowed a host of other priorities to undermine success in this essential endeavor. By taking the fully integrated testing of the weapons off the table requires that we bring our very best to everything else we do.
I’ve written a great deal about how bad our approach to modeling and simulation is, but it’s the tip of the proverbial iceberg of incompetence and steps that systematically undermine the work necessary to succeed. Where modeling and simulation gets a lot of misdirected resources the experimental and theoretical efforts at the Labs have been eviscerated. The impact of this evisceration on modeling and simulation is evident in issues with the actual credibility of simulation. This destruction has been done at the time when they are needed the most. Instead support for these essential scientific engines for progress have been “knee-capped”. Just as importantly a positive work environment has been absolutely annihilated by how the Lab’s are managed.
Without the big integrated experiment to tell you what you need to know for confidence all the other experiments need to be taken up a notch or two to fill in the gap. Instead we have created an environment where experimental science has been lobotomized and exists in an atmosphere of extreme caution that almost assures the lack of necessary results for healthy science. Hand in hand with a destruction of experimental science is the loss of any vibrancy of theoretical science. The necessary bond between experimental and theoretical science has been torn asunder. Usually when working well the two approaches push and pull each other to assure progress. With neither functioning, science grinds to a halt. Engineering is similarly dysfunctional. We do not know enough today to execute the mission. In a very real sense we will never know enough, but our growth of knowledge is completely dependent on a functioning engine of discovery powered primarily by experiment, but also theory. Without either functioning properly modeling and simulation is simply a recipe for over-confidence.
We can only see a short distance ahead, but we can see plenty there that needs to be done.
― Alan Turing
We have gotten to this point with the best of intentions, but the worst in performance and understanding of what it takes to be successful. We are not talking about malice on the part of our national leadership, which would be tantamount to treason, but rather the sort of incompetence that arises from the political chaos of the modern era. When we add a completely dysfunctional and spoiled public consciousness governed principally by fear we have the recipe for wholesale decline and the seemingly systematic destruction of formerly great institutions. Make no mistake, we are destroying our technical base as surely as our worst enemy would, but through our own inept management and internal discord.
Let’s start with the first nail in the coffin, the “Tiger teams” of the mid-1990’s. We decided to apply the same forces that have made nuclear power economically unviable to the National Labs (nuclear power has been made massively expensive through over regulation, and a legal environment which causes costs to explode through the time-integrated value of money). This isn’t actual safety, but rather an imposition of a massive paperwork and procedural burden on the Labs, which produces safety primarily by decreasing productivity to the level where nothing happens.
Science becomes so incremental that progress is glacial. You almost completely guarantee safety and in the process a complete lack of discovery. Experiments lose all their essence and utility in acting as a hedge against over-confidence by surprising us. Add the risk aversion we talk about below, and you have experimental science that does almost nothing. As a result we get very little for our experimental dollar, and allow ourselves to do almost nothing innovative or exciting. So yes, safety is really important, and we need to produce a safe working environment. This same environment must also be a productive place. The productivity gains that we have seen in the private world have been systematically undermined at the Labs, not just by safety, but two other drivers risk aversion and security.
Guaranteed security is another pox on the Labs. This pox is impacting society as a whole, but Labs suffer under another burden. We pay an immense tax on our lives by trying to defend ourselves from minuscule risks associated with terrorism. We have given up privacy as a society so that our security forces can find the scant number of terrorists who represent almost no actual risk to citizens. The security stance at the Labs is no different. We have almost no risk or danger of anything, yet we pay a huge price in terms of privacy, productivity and work environment to avoid vanishing small risks. Instead of producing Labs that are so fantastic that we constantly push back the barriers of knowledge and stay ahead of our enemies, we kill ourselves with security. We keep ourselves from communicating, producing work and collaborating effectively for virtually no true benefit aside from soothing irrational fear.
Finally we have a focus on accountability where we want to be guaranteed that no money be wasted at any time. Part of this is risk aversion where research that might not pan out and doesn’t get funded because not panning out is viewed as failure. Instead these failures are at the core of learning and growing. Failure is essential to learning and acquiring knowledge. Our accountability system is working to destroy the scientific method, the development of staff, and our ability to be the best. To some extent we account not because we need to, but because we can. Computers allow us to sub-divide our sources of money into ever-smaller bins along with everyone’s time, and effort. In the process we lose the crosscutting nature of the Lab’s science in the process. We get a destruction of multi-disciplinary science that is absolutely essential to doing the work of stewardship. Without multi-disciplinary science we will surely fail at this essential mission, and we are managing the Labs in a way that assures this outcome.
All of this is systematically crushing our workforce and its morale. In addition, we are failing to build the next generation of scientists and engineers with a level of quality necessary for the job. We are allowing the quality of the staff to degrade through the mismanagement of the entire enterprise at a National level. Without a commitment to real unequivocal success in the stewardship mission, the entire activity is simply an exercise in futility.
We seek guaranteed safety and that simply cannot happen without doing nothing at all. We seek guaranteed lack of risk, and no chance of failure, which is the antithesis of research and learning. Science is powered by risk and voyage into the unknown. Without the unknown, an inherently risky thing, science is simply a curating of existing knowledge. Our security stance seems totally rational especially in the post-911 world. It is nothing more than a fear mongering that strives to do the impossible, maintain a tight control over information based on science, and maintain our advantage by keeping us from using the best available technology. Instead of enhancing our productivity with technology and science, we hamstring ourselves to defend our possession of old knowledge. The power of the Labs and their staff is driven by achievement and discovery and the push for safety, risk-free and total security is completely at odds and work to destroy what used to be our strength.
The reasonable man adapts himself to the world: the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.
― George Bernard Shaw
When we look at the overall picture we see a system that is not working. We spend more than enough money on stockpile stewardship, but we spend the money foolishly. The money is being wasted on a whole bunch of things that have nothing to do with stewardship. Most of the resources are going into guaranteeing complete safety, complete absence of risk, complete security and complete accountability. It is a recipe for abject failure at the integrated job of safeguarding the Nation. We are failing in a monumental way while giving our country the picture of success. Of course the average American is so easily fooled because if they weren’t would our politics be so dysfunctional and dominated by fear-based appeals?
Evil people rely on the acquiescence of naive good people to allow them to continue with their evil.
― Stuart Aken
What could we be doing to make things better and step toward success?
The first thing we need a big audacious goals with enough resources and freedom to solve the problems. Stockpile stewardship itself should be enough of a challenge, and we do have the resources to solve the problem. What we are missing is the freedom to get the job done, and the general waste of resources on things that contribute nothing toward success. Actually much of our resourcing goes directly into things that detract from success. Think about it, we spend most of our precious money undermining any chance at succeeding. One of the core issues is that we are not answering the new questions that today’s World is asking. Instead we are continuing to try to answer yesterday’s questions even when they are no longer relevant.
Theories might inspire you, but experiments will advance you.
― Amit Kalantri
Another way of making progress is to renew our intent towards building truly World-class scientists at the Labs. We can do this by harnessing the Lab’s missions to do work that challenges the boundary of science. Today we are World class by definition and not through our actions. We can change this by simply addressing the challenges we have with a bold and aggressive research program. This will drive the professional development to heights that today’s current approach cannot match. Part of the key to developing people is to allow their work to be the engine of learning. For learning, failure and risk is key. Without failure we learn nothing, just recreate the success we already know about. World-class science is about learning new things and cannot happen without failure, and failure is not tolerated today. Without failure science does not work.
A big piece of today’s issues with the Labs are a deep disconnect between experiment, and theory that is necessary to drive science forward. As well as the admonitions against failure, the push and pull of experiment and theory has broken down. This tie must be re-established if scientific health and vitality is to be restored. When it works properly we see a competition between experimental science and theory. Sometimes experiments provide results that theory cannot explain driving theory forward. At other times theory makes predictions that experiments have to progress to measure and confirm. Today we simply work in a mode where we continually confirm existing theory, and fail to push either into the unknown. Science cannot progress under such conditions.
Much of the problem with the lack of progress can be traced to the enormous time, effort and resource that go into useless regulation, training and paperwork. These efforts go far beyond the necessary level of seriousness in assuring safety and security to trying to guarantee safety and security in all endeavors. These guarantees are foolish and lead to an overly cautious workplace where failure is ruled out by dictum and risks necessary for progress are avoided. This leads to lack of progress, meaning and excellence in science. It is a recipe for decline. We do not have a system that prioritizes productivity, progress and quality of work. We have lost the perspective in balancing our efforts in favor of the seemingly safest and securest mode of effort.
The stupid, naïve and unremittingly lazy thinking that permeate high performance computing aren’t just found there. It dominates the approach to stockpile stewardship. We are stewarding our nuclear weapons with a bunch of wishful thinking instead of awell-conceived and executed plan. We are in the process of systematically destroying the research excellence that has been the foundation of our National security. It is not malice, but rather societal incompetence that is leading us down this path. Increasingly, the faith in our current approach is dependent on the lack of reality of the whole nuclear weapons’ enterprise. They haven’t been used for 70 years and hopefully that lack of use will continue. If they are used we will be in a much different World if they are used, and a World we are not ready for any more. I seriously worry that our lack of seriousness and pervasive naivety about the stewardship mission will haunt us. If we have screwed this up, history will not be kind to us.
You have attributed conditions to villainy that simply result from stupidity.
― Robert A. Heinlein