The “Prediction-normalized Vote” and the “Mirror Budget”
From Mirror Budget to Weak Anarchism (5)
The “Mirror budget” may be integrated with the “prediction-normalized vote” mentioned earlier, from that point of voting（budget allocation） with an awareness of one’s position on the whole.
If “how to allocate the budget in the Mirror Budget” is thought of as “a vote for a particular option (i.e., budget allocation)” (or an individual’s perspective on the world), it can be mapped to the prediction-normalized vote by asking people to also enter “how much of a minority they think they are”.
Then, using the “predicted minority” degree, we weigh and tally the budget proposals and use that as the overall budget. However, the system can be manipulated to pretend to be a minority with no meaning, so we incentivize bonuses based on the accuracy of the “predicted gap” and the “actual gap” within some period to prevent cheats from reporting themselves as an over-represented minority.
This bonus is also decided so that “voting and predicting honestly would be most beneficial to that person”. The authors of the agreement voting paper have already provided a method for determining this (e.g., Measuring the Prevalence of Questionable Research Practices With Incentives for Truth-Telling. The authors offer a free PDF of the paper).
Furthermore, as the budget proposal of the “Mirror Budget” should be slightly different for each individual, it is necessary to calculate the similarity in some way (i.e., “embedding”), and various techniques related to machine learning are already well known for this.
It will rather be a challenge to make this embedding method fair and to keep its fairness from deteriorating. Because the algorithms for calculating budget similarity will affect “whether or not they are actually determined to be in the minority”.
Choosing several embedding algorithms at random, for example, to keep fairness, is a naive approach. In this example, “degradation” would be, for example, “ developers say it is random, but actually, plant a program for selecting methods that skew the results.
If the “Mirror Budget” and the “prediction-normalized vote can be successfully merged, the “Miller Budget” has the potential to be a budget that takes into account market outsiders by incorporating the perspectives of the “ordinary people with a long-term perspective, without locking the “elected officials with a long-term perspective” into the system as positions.
Now, suppose that the proposed budget derived from the Mirror Budget appears to be superior. Since the Mirror Budget is just a budget proposal, there is no compulsion to implement it. In that case, we would already have a concrete example of “immediately achievable reforms, i.e., budget details that are available, but simply not implemented.
Which is better than the current system of “someone who was chosen will figure out how to do it later, or after the system has been changed”? The strength of that persuasion and hope would depend on the level of detail.
Our hopeful observation is that the following situation is exposed. That is, while the detail of the Mirror Budget has the complexity and natural connection of a naturally occurring city, the current budget is so artificial and top-down that it resembles the awkwardness of a planned city.
Furthermore, if the actual government budget and the Mirror Budget with adequate participants significantly diverge, the question will inevitably arise, “Who is the government for in the first place? Because the Mirror Budget is detailed, unlike elections, where festive moods can be produced to get away with it, it can have the power to force an answer, “Why is it different from that?”