I am interested in the potential of automated reasoning and how this is done in groups. Since LessWrong has done such a good job of promoting and demonstrating feasible rationality, I am asking for insight as to how this could work with ‘dialogue’ or reasoned debate. I have a general concept in mind about how GitHub could be used and wondered if anyone could share their thoughts on how they see ‘version control’ been used to help people make better decisions and come to a concensus easier.
If templates could work well for certain topics, replicating and extending functionality would be easier and could become a standard operational perspective when discussing issues and solving problems. I realize I need a unique and novel angle to exploit and initially I want to explore the extent to which programming elements can be used to extend the functionality of adding and approving elements of the discussion. This would also include the evolution of the templates and naturally the topics they can handle for improving scope.
The problem I have, other than demonstrating the specific benefits of choosing GitHub as the mechanism to evolve ‘reasoned debate’, is that a certain methodology’s effectiveness and proven worth to reach consensus would have to be shown for each type of argumentation topic. I am not assuming this would be hard to do and would count towards an initial MVP of the project.
Any feedback or alternative perspectives worth considering and/or potential implementation issues would be highly appreciated. :)
I’m still not clear on what exactly you’re wanting to do with Github.
Can you give an example use case for your project?
What do you see the “templates” doing in this project?
Here’s an article (https://arstechnica.com/tech-policy/2018/11/how-i-changed-the-law-with-a-github-pull-request/?comments=1) on how Washington DC is using GitHub to update and maintain its laws. The suggestion from the article is that citizens would be able to make changes and take a more active involvement in the creation of laws. I’m not necessarily suggesting the possibility because there’s a number of strong reasons why this might not be a good idea (if you read the comments).
Could something be applied to collective reasoning?
The templates in this sense could be used as the format to reach consensus (like a law?). Let’s say a group is discussing a political topic and all parties involved have mutually agreed to a number of objectives of the dialogue. Including mutual respect for differing opinions and the need to upheld rigor and principles to maximize the chances of all agreeing and having the optimal outcome. In this sense, prior to the discussion, there would be formats to follow to reach an agreement. So, depending on the topic and which appropriate template is chosen the chances of success are ‘almost’ guaranteed because the underlining logic is agreed upon and already proven.
Therefore, the question could be, is there a format of taking differing opinions (inputs) at certain stages of an argument, which if the evidence and results (output) are agreed upon can solve the initial topics question and then be applied to any number of topics (if in a certain format).
In this sense, you would be ‘coding’ or adding to the original document your position and reasoning of certain subsets of the overarching logic of the argument. These would be agreed upon prior to when the template is chosen. Meaning you could complete a number of reasoning practices before the different parties are actually engaged in the mental activities of evaluating judgment and critique etc (arguing).
Using version control for law making doesn’t change much about citizens ability to suggest changes for laws. What it does do is that it makes it easier to track who proposed a change. If a lobbyist for example sends a congressman a pull request for a law it’s a lot easier to understand and track how the law came about. Version control increases transparency and it also has the prospect of making law making more efficient as it’s easier to keep track of amendments that way then by mailing Word documents around which is the status quo.
The process of how the law was changed in that instance isn’t any different then how you can change the code of an Open Source project. I know of no open source project that’s using version control for the discussions about what changes to make and having discussion to find consensus.
Instead of looking at version control it would make more sense to look at software that’s actually designed for consensus finding. Various liquid democracy software provides for consensus processes. Taiwan’s e-government initiative is also worth checking out.
Here’s it’s worth noting that judgement and critique is not the necessarily the rational way to come to consensus. If you look at the processes CFAR developed seeking to understand another, identify cruxes and explaining why you hold your position can be a lot better.
There are a few things that sound similar to what you’re talking about. The first is the process of writing an RFC: https://github.com/inasafe/inasafe/wiki/How-to-write-an-RFC. Also wikipedia must need to do many of the things you describe, so looking into how they reach consensus may be interesting for you. Also, there are attempts to have more of a direct democracy style governance in the US, and they have certain procedures that you may want to look into: https://www.newyorker.com/news/the-future-of-democracy/politics-without-politicians
I do like the idea of templates for certain types of discussion. That’s why I wrote this: https://www.lesswrong.com/posts/xE7F4b34pfTMThYMX/what-questions-should-we-ask-ourselves-when-trying-to.
It’s not easy for me to understand how strong of a claim you’re making here, because you say “depending on the topic” and “almost.” It still feels too strong to me. I’d say most of the time, at best, templates for discussion would just be helpful. Especially if people have different values and beliefs about the world, disagreements are very difficult to settle.
I suppose questions in mathematics or something where you can prove an answer is correct may be a type of exception. Check out the polymath project if you haven’t seen it already for an example of people collaborating on trying to solve (math) problems.
I have a lot of similar ideas to the ones you’ve presented in this post, so if you’d like to discuss these things anytime, feel free to send me a dm.