Dedomenocracy: separation of science and state

In this blog, I have contemplated the particular concept of government that I termed a dedomenocracy that imposes short lived policies only during times of urgency and those policies would be automatically computed from some pre-approved algorithm working on the totality of data that is available. It is a futuristic government because it must have access to abundant data, and as close to exhaustive as possible. We are close to having the technology and infrastructure to implement such a government.

There are multiple ways to implement a government by data, but I am focused on a very narrow definition that eliminates any human involvement in the selection of the policies. No human can overrule or modify the selected policies. Once selected, the population is obligated to follow the policies and there would be authoritarian enforcement without regard to the specific circumstances of the individuals. The policies inevitably would impose greater burdens on some parts of the population and possibly grant greater benefits to other parts of the populations.

The unfairness of the policies are mitigated by the fact that the population has control over triggering any new policies, and every new policy has a specific expiration date that is at most a couple years into the future. The population triggers the policies by a super-majority expression of some urgency. This immediately provides a safeguard because few crises will raise to the level of a super-majority request for action. Also, the conditioning from past policies would serve as a deterrent to join the petition for action. In particular, although the population expresses some urgency, the resulting policy determined by algorithms and data may be unfathomably related to the urgency and would impose burdens and rewards seemingly randomly. There is no way to predict how any person would fare with the algorithmically selected policy.

In my view, the population has control over the collection of data and of assigning grades to various sources in terms of trustworthiness, applicability, or conformability. In a dedomenocracy, the people are involved in government, but in contrast to democracies where people select human leaders, they instead select the data to allow the algorithms to consider.

The population also has a role in selecting and tuning the algorithms. In particular, the population would decide the objectives whether they are to take advantage of beneficial opportunities, or to avoid potential hazards. The population would also decide the relative priorities of competing objectives based on confidence levels given the data at the time. In my imagined implementation, these algorithms would be relatively stable over time and certainly be in place long before an urgency would come up.

This concept of government provides an advantage over democracies because when a crisis comes up, the population has confidence of how the merits of various objects are computed and prioritized. In a democracy, we rely on humans who may decide to act completely differently than how they campaigned. Democracy selects human leaders who are permitted to react based on their individual judgements or emotions that definitely will change depending on the nature of the emergency. In broad terms, a dedomenocracy will tend to emphasize long-term goals while a democracy will almost exclusively concentrate on short term goals, usually in the form of maintaining the preexisting order.

Within my discussions of this government, I spent a lot of time discussing different types of data in terms of their appropriateness for decision making algorithms. In particular, I graded data by a scale I call brightness where the brightest data corresponds to direct and recent measurements from trusted and well calibrated sensors. At the other end of the scale is what I call dark data that comes from computer models instead of direct observations. The deliberate connotation is that brighter data is preferred over dimmer data and the least preferred is dark data.

Dark data is data from computer models. Computer models are typically based on science. As a result, my concept prefers sensor data over scientific theories. This comes from a trust in the algorithm’s ability to infer the science from the observed data. That trust comes from observations of the effectiveness of self-learning artificial intelligence that derive their own models of how the world operates instead of having some scientist pre-program some theory. Also, the trust presumes that the algorithms will be sufficiently powerful enough to ingest an exhaustive amount of data and that the sensors are capable of providing that exhaustive data store. With enough observations and training data, the algorithms ought to be able to come up with their own models of how the world works. Those models will be inaccessible to scientists for comparison with human created models. Attempts to feed human created models into these algorithms will discourage the algorithm from discovering some new fact of reality that science has not yet discovered. In particular the algorithm can find relationships between a multitude of variables, far beyond what is possible from human comprehension.

Summarizing the above, my concept of a dedomenocracy excludes human science from decision making. Decision making is based entirely on algorithms deriving some intelligence based on all of the available observed data, and with an emphasis on the most recently observed data. By implication there is a separation of science from government.

One of the principles of our current government is that there must be a separation of church and state, there is no state sponsored or mandated religion. A trivial interpretation of this separation is that the government cannot tell any person what religion to follow or to obey. More broadly, this separation demands that government find some other justification for laws other than religious doctrine. Governments can not justify their legitimacy on religious arguments, and they can not legitimize their laws by appealing to religious documents.

The typical solution to this prohibition is to justify policy based on science instead of religion. Laws must be defended by some form of scientific argument or proof.

This approach worked very well in the earlier histories of modern democracies. During this earlier period, there was a broad and diverse practice of science by the general population. People from a wide variety of backgrounds, training, and occupations would participate in science, either by adding to science understanding through their own investigations, or by challenging other theories through peer review and debate. Early on, most if not all of the scientists operated independently of government funding or approval. The population would present the science to the government, and the government’s role was to be persuaded by this externally offered science.

Over recent decades, and certainly all of my lifetime, the government has become the primary funder and approver of science. Currently, the vast majority of people doing recognized science work directly under government paychecks, contracts, or grants. These scientists’ ability to continue doing science depends on the government’s confidence that the scientists will deliver science that the government can approve.

Increasingly over time, the government constrained the type of inquiries it was willing to fund. Such constraints offer some economic efficiency by no longer challenging already accepted science. There arose a concept of settled science, areas that no longer need funding. This permits the funding to pursue areas that are not as well settled.

I argue that this concept of settled science is a recent development. In the past, when science was more broadly supported through sources outside of government, there would always be the opportunity for someone to challenge any aspect of current understandings. They would be able to present their findings to the respective scientific community and that community may adopt even major changes to theories when the proof was compelling enough. This was possible because there was no single source of funding of science.

The current government monopoly over scientists goes further to constrain its funded scientists by prohibiting them from discussing challenges to the officially approved policy that was based on some science. This is especially true for scientists directly employed by government or by its contractors. While there can be internal debate within particular departments and agencies, there is a prohibition of exposing the detailed compelling arguments outside of those departments and especially against exposing these challenges to the general public.

When employees or contractors desire to participate in outside conferences or publications they are obliged to have their work reviewed by public affairs and other arbiters who would check the work for any information that is not allowed to leak out of government. A large category of information prohibited from disclosure involves information that can undermine the public’s trust on established policies.

In recent months, it is remarkable how consistent are the scientific descriptions of the problems and the solutions. Even as the understanding changes over time, all voices of the scientific community express consistent viewpoints immediately. The evidence of the changes in recommendations hints that there are debates going on. The public does not hear of the details of the debates. The public is not permitted to participate in this debate.

I’ve seen cases where someone outside of government funding would express some contrary theory and then be immediately silenced for spreading misinformation, and then later learning the government adopting the exact same theory. When the government accepts the theory, the announcement justifies the change as being consistent with a scientific method.

The scientific method is exclusively within the confines of government control. Outside participation is not allowed, especially if that participation challenges the government’s policies.

Throughout this, the government is avoiding the government by religion doctrine by instead justifying its decisions on science that happens to be science that the government completely controls. This is effectively a violation of the separation of church and state but where government-controlled science substitutes for the church. The church of science excludes from its priesthood anyone who is not employed by the state to do science.

Recent policies for managing the current crisis are remarkable because of the lack of publicly expressed dissenting opinions. These are very severe and disruptive policies and yet we are told that science completely supports these policies. Many different governments adopt identical policies and this further emphasizes the settled nature of the science.

Meanwhile, there are outsiders pointing out flaws or gaps in the scientific basis for these policies or recommendations. These dissenters are almost universally denounced as being anti-science or as state enemies guilty of spreading deliberately misleading information. Many of these dissenters make very compelling arguments to the general public, many of whom also have some scientific education and training even if their jobs do not reflect that.

The remarkable aspect is that there is so much silence from within the government funded scientists. Apparently not one of these government funded scientist find the dissenting arguments to be convincing. This is very unusual because it is contrary to human nature. Naturally there must be some people within government who have similar dissenting views. They are not expressing their dissent to the public, and probably not even within their own departments.

The government hires a very large number of people with scientific training because masters degrees and PhDs are requirements for many positions. These people have the skills and aptitudes to come to their own conclusions about available data. That data is sufficiently ambiguous to permit multiple interpretations. Yet, we hear only one interpretation ever coming from anyone who has claim to being an expert.

Having a financial relationship with the government restrains a person’s ability to engage in the science.

For those with specific job roles, they are required to spend their hours on their assignments. Any participation outside of their assigned areas distracts from the work they are supposed to be doing. Thus even if they recognize some flaw in scientific reasoning in a different agency’s policies, they are not permitted to invest in the effort to make their case. They are silenced due to compartmentalization of duties.

When the person does work within the relevant agency, there are still limitations imposed by job duties. Someone working on maintaining computers in an organization basically has no say with the analysts making the policies. Their work must remain within their job description and pay grade.

Within the particular agency or department, there is the specific group of analysts whose work directly supports the science behind the policies. These analysts are constrained in a different way. Typically, their tasking is to develop the scientific justification for continuation of some existing policy, or to develop new justification for some new policy. The process is backwards. The scientist task is to bolster the policy. Government scientists rarely are given the task to recommend a policy from current science. This is especially true after the policy is already in place.

Though never stated as such, there is fundamental conflict of interest among modern scientists. They all depend on their good reputations with the government in order to continue to jobs and salary. They need to provide science that satisfies the government, and they to avoid any science that can undermine the government’s position.

We are conditioned to think that conflict of interest does not apply to government scientists. Being employed by the government shields the scientists from any conflict of interest. This is fundamentally not true when the government uses the scientist’s work to decide and reinforce government policy. All government funded scientists, whether through salary, contract, or grant, have a conflict of interest when it comes to providing science to support government policies. The strong bias is toward supporting those policies.

The conflict of interest within government scientists is equivalent to the accusations of conflicts of interests of industry funded scientists such as the archetypical tobacco scientists showing studies that tobacco is not addictive. In both cases, the scientists have a large financial incentive to please the ones providing their funding. That same incentive strongly discourages them from exposing anything that would displease their funders.

Back to my discussion of dedomenocracy and its demotion of dark data. Dark data is data coming from models based on human science. The conflict of interest is avoided by making decisions with preference on observations instead of scientific models. The actual immediate policy, when it is triggered by some urgency, is not dependent on science. The policy is dependent on observations. The assumption is that the algorithm has access to abundant and nearly exhaustive observations. Those observations come from trusted sensors collecting real world data. That data is constrained by the laws of nature of the real world. Those observations can rederive the science if that is appropriate to the circumstances. More importantly, the observations are free to derive some new science if the circumstances support it. Dedomenocracy’s derived new science is free to contradict the old science if the data supports a new understanding.

With the current situation, there is a lot about the observations that show that the epidemic and the policies are not behaving anything close to what was predicted. The government is committed to its policies, and the government-funded scientists are committed to support those policies. What started as a two week emergency has stretched to over a year with predictions that it may last another year or two. Meanwhile, the observations are strongly suggesting that there may be better theories and thus policies to deal with the circumstances. The current government and its grip on scientists does not permit the necessary deviation to take advantage of this new information. The governments are committed to follow the one and true science that comes from government funded scientists who the government allows to speak to the public.

The governments are in effect adopting a state religion, and that religion is disguised by a different name of science. This government religion has its government sponsored priests whose pronouncements determine the policy. The public has no recourse to challenge these pronouncements because they are not government approved and funded scientists. The public can not do science on their own.

The transformation of science into a religion occurred relatively recently although before my time. Earlier there was a more popular interpretation of science. Science and critical thinking were skills that everyone was encouraged to acquire. Anyone acquiring these skills could put those those skills to work to prepare arguments to challenge some existing science or to propose some new science. They would present their arguments to their peers in the sense that everyone was in roughly the same status in terms of their having independent sources of funding, often being self-funded. The arguments were considered on the merits instead of the scientist’s job positions. Certainly, there were a few highly respected scientists whose opinions demanded particular respect. However, generally, anyone could participate in the science and have the opportunity to make their own contributions, just as many actually did during the 17th and 18th centuries.

We don’t have that today. I do not think we are better served with the current model.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s