Tauchain may allow for the development of the first "Smart Constitution"

in tauchain •  5 months ago

The concept of a smart constitution is something I've been contemplating since 2016. The idea being that you require a self amending layer to the protocol similar to what we have with Tezos but in order for it to be truly smart it has to go farther than this. Specifically Nick Szabo is critical of the "human interpreted wet code " style constitution of EOS and for very good reason; I'm now inspired to discuss the concept of a smart constitution which is inspired directly from Nick Szabo's idea of a smart contract.

The Smart Constitution concept for implementation on Tauchain

What will make Tauchain unique is the fact that contradiction and many of the problems associated with human interpreted wet code can be solved. One of the main issues of debate is on the amorality of the law in human interpreted wet code form which I discussed under the title: "Legal exegesis, the law as amoral, and the justice institution as consequence distribution mechanism". In addition to my discussion is the blogpost from Ohad Asor titled: "Art of Self Reference". In my post I specifically mention legal exegesis and it is important to note that exegesis was a word specifically chosen to highlight the manner in which law is interpreted (it is interpreted in a similar manner to that which religious scripture is interpreted).

This subjective interpretation of the law and constitution is why we have higher courts. The Supreme Court is the final arbitrator of matters involving interpretation of the constitution or the law. Who exactly would act as this function for something like EOS? The problem putting a traditional wet code style constitution into EOS is that now it creates a market from which will emerge a priest class. This priest class will function to interpret the EOS constitution so that we all can agree on what it means. If there is no priest class then concepts such as harassment have to be very narrowly defined with zero ambiguity. We know that the English language and natural language in general includes ambiguity as a feature (but also a curse) and because of this lack of precision there will for sure be disputes as to the meaning of certain words within the EOS constitution, semantic arguments, conceptual arguments, who will resolve these disputes fairly?

EOS also risks recreating a caste system (and I was concerned about this when the matter of privacy vs radical transparency was discussed). What I mean by caste system is you will have the priest roles which will likely be a requirement. You will have politicians (witnesses) who basically have to get votes creating a political role. You will have to also enforce this constitution so will it be self enforcing in code, or legally enforced, or enforced via coordinated shunning campaigns? How will rules be enforced is another question to ask.

What makes Tauchain particularly unique is the fact that many of the problems EOS will face (including the risk of recreating a caste system around moral hierarchy) is that on a fundamental level Tauchain is supposed to be designed to help people with good intentions to become better. If you use Tauchain with a sense of morality then using Tauchain itself you will be able to receive answers to your moral questions. The other unique feature of Tauchain is the fact that there will be no question about how to interpret the rules because the logic and deductive reasoning will be built into the platform. Tauchain will capture our worldviews, our morality, our knowledge, and when we create rules or mutually agree it will be in ways which do not contradict our own interests because of the logic being a built in feature.

So for example deontic logic could be utilized so that the rules can be formulated without ambiguity of natural language. Something is either permitted or it isn't. Something is either obligated or it isn't. This logic can be built into the code itself and discussion about the code can scale so that a lot of the problems which may be inevitable with EOS can be avoided with Tauchain.

So what is a smart constitution exactly? A smart constitution is a concept which Tauchain by it's design will enable. That is a continuously upgraded set of rules which form the basis of how Tauchain is to operate. These rules can be set up in such a way that new future rules can be added which never can contradict these original rules. So it will be very important in the very beginning to care about which rules really are essential to operating the network and which rules are not. From the most essential there will be other rules which derive from these rules.

For example an essential rule could be that Tauchain as a platform cannot be shut down unless certain conditions are met. These conditions could be anything from a threshold of votes or it is also possible that the owners of Tauchain could decide no matter what it should never shut down. If it never is to shut down then any operation which threatens to shut down or damage the ability of the network to keep running would potentially violate this rule.

Another example of an essential rule is that Tauchain most provide some social utility in order to receive resources. This kind of rule would be more difficult because utility is something not objectively measured but we do know that if enough people think Tauchain is somehow harmful then they will not want to contribute resources to the network. The point here is the network itself could be designed to evolve and adapt in such a way so as to always seek to provide a measurable benefit to mankind.

Many potential rules could be formulated or could exist. The point being that a smart constitution would take into consideration the morality of humanity, the knowledge of humanity, and create a sort of consensus. This consensus would govern how Tauchain itself can evolve, how resources of Tauchain can be managed, etc. What would make the Tauchain "smart" constitution smart? The fact that the Tauchain platform itself will help us to improve the constitution based on our preferences and interests.

This topic is: To Be Continued. One post will not be enough to explore the possibilities.

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

I've also dreamt about a system like you describe for quite some time. The benefits to mankind could be enormous. The range from humanitarian and social problems that could be solved in an agreeable manner is quite astounding. I am very curious how this will work out in a used network, specifically if the rules set in a smart constitution could be valid and apply to resources outside of the defined system (tau/the application running on tau). I am also very interested on how a user-interface for up-scaling discussion might look, and how it would keep peers w/l little time or knowledge on subject engaged.

The beauty of it is, that these decisions could also be made using up the inherent logic of the system. Your post made me a bit more excited about the future again, thank you!

Well Tauchain certainly isn't the first and won't be the last attempt to formalize governance on the blockchain itself - anyone remember Tezos? Yup Tezos is actually coming back as T2 sometime this year. Last I checked Tauchain also had some issues too.

And remember, we had the DAO which wasn't a bockchain with its own governance encoded on chain, but an org on a chain (Ethereum) with its own governance in contracts. Problem was they could never get enough votes to decide anything and DAO became a quagmire of inactivity. Until the famous hack that is which drained it of millions in funds and famously sparked the Ethereum Classic and Ethereum fork forcing the human nature of Ethereum governance to front and center stage.

The problem seems to me, that law as code is fine if you can describe all the things you want to legislate as code. Think of the US constitution, and then think of the thousands, if not millions of laws that have been created underneath it to make it actually a practical system. How do you decide if one law in code is a legitimate law - a subclass if you will - of the original "constitution" that the system started with? Take for instance the first amendment - seems pretty clear right:

Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.

But somewhere along the line we figured out there are limits to rights in the Bill of Rights - such at commercial speech is limited, obscenity is limited, yelling "fire" is limited, incitement to specific criminal acts is limited etc. and I would think most people would agree those are useful things. But however would you have encoded that in code?

You note the ambiguity of English language for expressing laws (gun rights advocates will spend hours arguing about the meaning of a comma in the second amendment because it suits their interpretation) but you didn't mention when the effect or cause that is to be limited is a very human thing. We all know that famous quote about obscenity from a legal case Jacobellis v Ohio:

I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description ["hard-core pornography"], and perhaps I could never succeed in intelligibly doing so. But I know it when I see it, and the motion picture involved in this case is not that.

So what are we to do when a human can't even describe the thing and it takes a human to determine what is or isn't obscenity? Assuming people agree such things are necessary laws then how do we codify them? Do we train up a machine learning algorithm to get a high F1 score for determining obscenity and make that part of the blockchain? Or can we always call out to human-driven oracles who provide input to governance?

When your neighbor is having a loud party do we require a law that has expressly defined every environmental parameter of what defines a nuisance and have an Oracle feed all that data into the blockchain? Or do we just use an Oracle that is a front for human judges to resolve disputes - something like Kleros which offers a dispute resolution protocol and applies game theory to encourage good judging.

Whatever you think or happens I have to agree that blockchain governance is a fascinating field and arguably the most important one to be addressed and with the biggest potential for impact on human society. I honestly expect that in future years - maybe 50 years hence - we will all be living out our lives at the intersection of many blockchain governance strategies and fields. Everyone will be able to immediately asses what governance rules apply to anyone or any entity (including AIs) they interact with and act and transact appropriately. We might ever well end up with a constitutional convention that basically delegates pretty much everything to blockchain regulated governance. Of course people wishing to live like that will necessarily need to instrument the heck out of their lives - but by then no one will think twice about that and that could very well end up looking like living in an episode of Black Mirror as is already been implemented in China with their social credit system augmented by mass surveilance and monitoring of all communication: http://www.businessinsider.com/china-social-credit-system-punishments-and-rewards-explained-2018-4

·

I hold Tezos so of course I remember. Tezos has the opportunity to learn from the mistakes of other platforms like EOS or Ethereum. Tezos being self amending at least has the main component necessary for a smart constitution but it's not enough by itself. You need also the ability to create rules which are consistent, which don't contradict.

·
·

Is there a specific mistake of EOS you're thinking of? Or is it just a lack of on-chain governance? It is my understanding that the EOS constitution could be amended to include that in the future. As it is with EOS roughly following the Steemit model I think we have something that is a happy compromise between a completely automated on-chain system, the risks of a susceptible benevolent dictator in Ethereum, and economic and environmental disaster of miners uber-alles Bitcoin. I guess we'll just have to see how the community-blockproducer-constitution system works out.

I don't follow the governance side of Steemit too closely but I feel like it is reasonably successful - other than the internal economics that could skew how the message gets out when we need to vote for producers, plus the ability to game the system with multiple accounts.

Do Tezos or Tau do anything to try and create strong identity?

·

On the topic of filtering speech this in my opinion is what collaborative filtering can best handle. Filtering basically isolates the noise (what you would subjectively interpret as obscenity) so that it is not brought to your attention. So yes you can filter just fine and probably far more effectively with something like Tauchain because you could unambiguously and precisely define exactly what you consider to be the noise vs what you consider to be most valuable (and prioritize accordingly).

The point is you will not have to worry about contradicting yourself, or for example the situation where a person claims they believe in absolute free speech under specific conditions yet they don't because they don't vote in that manner? If you believe in absolute free speech then what is obscenity? Obscenity is the speech you don't want to see, the information you don't want to consume, the spam. Ideally no one should be forced to consume or see information they consider to be noise.

References

  1. https://en.wikipedia.org/wiki/Collaborative_filtering
·
·

Collaborative filtering makes a lot of sense but carries a risk with its propensity for creating bubbles, isolation, and echo chambers. Filtering out everything you don't like can make for a peaceful experience, but in as much as online life is now strongly influencing the real work - to the point that it definitely affects elections now - it leads down that path towards one of those Black Mirror dystopia.

People have a propensity to be afraid or wary of things they haven't seen before. It's a vestigial survival skill that manifests itself in xenophobia and irrational dislike of things we have no experience of. I'm sure if you moved someone from 100 years ago to present time and plonked them down in a big City they would freak out at what they see in terms of diversity, dress, social norms etc. But given time I reckon they would adapt - just like someone coming out of the Amazon or a remote tribal life. However if you just shut out the rest of the world you're never going to change. You're never going to think gay marriage is okay, or back in the day that women or non-whites should get the vote, you'd never hear or see their voices and opinions. You'd never know that a clear majority of Americans think marijuana should be legal. You'd probably never know anyone that had come out. You could be just stuck in your community filtered bubble of fake news and very narrow opinions that would almost always agree with yours.

That's what I'd worry about...

·
·
·

Filtering out everything you don't like can make for a peaceful experience, but in as much as online life is now strongly influencing the real work - to the point that it definitely affects elections now - it leads down that path towards one of those Black Mirror dystopia.

Dystopia is caused by other problems which I address. It's not enough to give every human a vote if people aren't allowed to be wise. Social media doesn't encourage wisdom building, or good decision making, or knowledge diffusion. Social media encourages people to make critical decisions such as who to vote for based on stuff like party loyalty, or conspiracy theories, or gut feelings, so we don't have a situation where people even vote in their self interest.

People have a propensity to be afraid or wary of things they haven't seen before.

Risk literacy tends to be low in people. People cannot assess risk very well but then are asked to make important decisions involving risk. Making decisions under uncertainty is about analysis of statistics. We aren't all going to be experts at that but maybe on a platform we will want to follow the people who are.

However if you just shut out the rest of the world you're never going to change. You're never going to think gay marriage is okay, or back in the day that women or non-whites should get the vote, you'd never hear or see their voices and opinions.

When faced with newer and better knowledge the correct thing to do is to update your worldview. This is a simple rule which could be implemented (if you think of it like a rule). This is called belief revision in the literature.

The point being that humans are too limited in ability to analyze information. Information is all around us and goes to waste. We also work with inaccurate models of reality which functions to keep us from acquiring or building wisdom. If wisdom building is as important as wealth building then we should want to build it but there isn't any reward for being wiser.

In other words ignorance is rewarded often and the forces of behaviorism are at play. The positive reinforcement, the intermittent reinforcement, all are involved at locking us into inaccurate models of reality and keeping us involuntarily ignorant. So of course when it's time to make critically important decisions we decide like we would a beauty pageant, or any other superficial contest.

Collaborative filtering is really the only way. The best we can do is make each individual a better thinker and decision maker if they choose to strive toward it.

References

  1. https://en.wikipedia.org/wiki/Belief_revision
Loading...
·
·
·
·

I don't see how collaborative filtering fixes any of the issues you and I mention with social media. If I create an account on Netflix and select even one conspiracy theory or junk science documentary I'll be flooded with recommendations for like material my supposed peer group has collaboratively filtered for me. And the more I watch the more I will find. I would posit that life gravitates to local minima surrounded by inviting does lying the unselective in. These insular swamps stagnate and self enrich creating their own ecosystem and cultures. The human propensity for social activity makes it hard to escape - I've talked to many former members of religious cults (which is about all of them IMO) and they relate that first hand.

The current culture of treating knowledge and critical thinking is a depressing trend and stinks of rank centralised control. It doesn't give me much hope can shake up the playing field and and bust out some people from their festering puddles, pools, and swamps of circular and unreasonable thinking.

Loading...
·

In my opinion a way for filtering to work is simple. Collaborative filtering (up and down votes, etc) combined with stigmergy (follow the voters who share your values). This will allow you to see more of what you want to see and less of what you do not. This will also allow you to direct resources toward what aligns with your values at all times (this can even be a rule).

This post is exciting and scaring at the same time. Many cultures have completely different views of morality. Still, it would get rid of a lot corruption. I'm always skeptical about relying on computers to judge humans...but I can also see the benifits of putting laws on a smart contract. Perhaps a hybrid of some sort. Anyway, great post as always. Thanks for sharing.

·

Computers don't judge humans. Computers (spreadsheets) help humans to judge humans.

·
·

Yes, but my fear is humans would allow whatever data was spit out of the computer to determine the future of a human, without any other consideration for particular circumstances. As long as we only use it as a tool, I can see this being very useful...although it does create a possible slippery slope.

·
·
·

Explain what you mean? How can we even agree on a definition of what human is without using computers? Can anyone be said to know 100% of what a human is? I would say probably very few if any.

So the concept and model of what a human is has to be defined. We have to then agree on what does and does not constitute a human. Once we all reach an agreement on what "human" is, then we can have discussions about what the future of "human" should be.

In my own opinion, most people don't know what human even is. Most do not have full understanding of biology, anatomy, or any of the necessary knowledge to create an accurate mental model. For this reason we have certain people in our world who deliberately dehumanize other people. How can we become more human if most of us don't even know how to define human and even if we agree on what human is then the question on what human can be is an open question.

For example is human a species permanently restricted to living on earth or does human as a species seek to evolve to exist on other planets? If the answer to this question is the humans want to evolve to live on other planets then the nature of what human can be is permanently and necessarily changed by that goal.

Tauchain is necessary to even have that kind of discussion. A person can on Steem claim they believe in space travel while also claiming they don't believe in transhumanism. They would be potentially contradicting themselves. A human simply cannot even travel in space without becoming what some call a cyborg (wearing a space suit, using machines to support life, etc).

So human in the unmodified traditional definition is as limited as the other species of earth.

Now as to using the data spit out of a computer to decide? That is exactly what a program is. A program decides. So for example if it's medicine then how do you think we determine kidney health? There is no direct way to measure it so we use a formula which runs on a computer using detectable enzymes in the blood like creatinine level. We basically take the individual human in the case of personalized medicine and we digitize them, and then we run tests or crunch on that digital representation.

The same is true for genome sequencing. If you want to know what a human is then you have to analyze the code which makes a human up. If you want to know what you are don't you think having your genes analyzed would give you greater insight into that? Of course you would learn from that and the question is would you let the computer tell you that you're genetically vulnerable and adapt to this or would you ignore it?

The computer can't make your decision for you. You ultimately have to decide. The computer and "thinking machines" only can help you to better decide by allowing you to export your thinking outside of your brain and into mechanical minds. Excuse me if I have no better way of expressing these ideas because this is next generation blockchain tech we are contemplating.

·
·
·
·

I think I understand what your saying...and thank you for taking the time.

In my opinion, you might be confusing what a human is with what a human does, or is capable of doing.
To keep it simple, a human is what is birthed from combining a male human sperm with a female human egg. Now we can choose to travel into space, or maybe even one day live on another planet. If we did, we may evolve to that environment, much like astronauts start to evolve when they spend to much time in space. Their bones get brittle, but their senses learn how to become oriented in a zero gravity environment. That doesn't change who they are or make then less human, it just means they've adapted to their environment. Humans have always been able to adapt.
To use an example on earth, when a blind person develops an incredible sense of hearing and smell to compensate for their lack of vision. That doesn't make them any less human, they've just adapted or evolved. So I think we can define what a human being is, while still allowing for changes. After all, no 2 human is exactly the same, but we are all part of the same species. Even if we start to incorporated computers to our body...which we already do with pacemakers and other medical devises, we still remain human. We're just attaching something to our body.

My only point in all this was to say we need to always make sure a human has the last word. I don't ever want to live in a society where computers judge us. We can, and do, use computers as tools to help us make decisions, but that's very different than a computer having the last word. Hope I've explained my self properly.

Wow, I know not of this before . Oh IMO, I can refer to tauchin as the Blockchain National Assembly. This is cool . Thanks for impacting knowledge . You’re much appreciated. I’d definitely learn to read more so I look forward to the exploration possibilities.

This is most interesting. It seems intuitively doubtful if it will be possible to formulate a constitution to rule out the need for human interpretation. I suspect it will be completely impossible. With artificial intelligence, it could be possible but it would require the kind of advanced AI that does not exist yet. As an experiment Tauchain has much potential to teach us valuable lessons. If Tauchain still exists, say, five years from now, it might be possible to use AI as the court system at least in some limited by important ways.

·

It's a language thing. There are languages which are unambiguous. There even are versions of English designed for reduced ambiguity. It's just that the language used in most law and in constitutions is not the sort of unambiguous English.

Learn about Lojban as an example of unambiquous language. There are others too. TauML translates from one language to another.

In 1987 the Logical Language Group began constructing a language based entirely on mathematical logic. As a foundation for their work, they used James Cooke Brown’s research from 1955. Brown had created a language called Loglan in order to test the effects of language on the speaker’s thought. The LLG adopted many of Loglan’s concepts to create their own language. Their goal was to invent a language that would be able to express complex ideas simply and without ambiguity. They aspired to remove the restrictions that ambiguity imposed on creativity, thought, and communication. By 1998, they had created an entirely new language according to those precepts: they named it Lojban. That same year they published a complete grammar and vocabulary of Lojban under the title The Complete Lojban Language.

Learn Lojban.

TauML would allow from simple English to Lojban. The problem would be greatly diminished even using natural languages.


·
·

Thank you for a thorough answer!

The problem with approaching modeling reality using a precise and unambiguous non-leaking abstractions is that reality is darn complex. That primarily arises from the properties of reality itself and not the language we use. There is no going around the fact that modeling the intricacies and complexities of real-world situations is a daunting task. It takes a human brain 20-25 years to reach the capability to function independently in the modern world.

Consider the task of developing a program capable of beating the best human players at the very well-defined task of playing go. It is utterly beyond human capabilities to express that in explicit rules. AlphaGo, the first program to accomplish that feat was based on neural nets. Now, a large neural net is computationally equivalent to a very complex function that takes a board position as its input and gives the move it considers best as the output. The AlphaZero network, a successor of AlphaGo has 84 layers and thousands of nodes. It took a vast computational effort to train the network and it takes a supercomputer to use it at a level sufficient to beat the best professionals. And go is a simple and well-defined abstract complete information board game!

Making the language as internally consistent as possible is but scratching the surface of the complexity involved in interpreting a constitution governing any collaborative effort. The interpretation of said constitution requires understanding the intricacies of every type of situation that may arise. If that could be done algorithmically, then we'd have created an artificial general intelligence (AGI). Nobody knows how to create an artificial general intelligence at the moment.

·
·
·

The role of Agoras in all of this is what happens when we are using TauML to translate so we can reduce ambiguity in our communication with each other, so we can reach agreement on what to do next? Well if we do reach agreement we translate it from human readable to machine readable, so from controlled English for example to machines via program synthesis.

So we can think of our agreements on what to do next as collaboratively describing the functioning of software in as detailed of a manner as possible (a specification). From this we either can rely on program synthesis in which case a directly translation into software or in other instances it will be human programmers who will need to be hired.

So when we have agreed as a community then program synthesis turns that agreement into code automagically. When one of us wants to do our own project or if paths diverge then Agoras will allow anyone who has tokens to communicate their specification to the network and then any programmer(s) can meet the requirements. In other words, a market for code paid for with Agoras.

And no not just code, knowledge, computation resources, etc. So you can see Agoras will be a big deal. As big as EOS by itself, but collaboratively developed and improved upon continuously. It is also true that Tauchain will potentially via Agoras create lots of jobs for digital nomads who will be able to sell knowledge, or do coding, or do knowledge engineering.

Loading...

To listen to the audio version of this article click on the play image.

Brought to you by @tts. If you find it useful please consider upvoting this reply.

Very interesting. This would need to be tested and retested before implemented...but it's certainly worth exploring. Great post @dana-edwards.

I must admit...this is a little scary. Great post, but kind of disturbing to think where it might lead. Very Orwellian in my opinion.

·

What about it scares you? Are you aware you post on Steem which has similar goals of facilitating discussion? The distinction is that Steem doesn't improve the quality of discussion but only allows for scaling the quantity.

Since you mentioned tauchain i have to look into it now. Cause i am still new with Eos, i haven't invested in it but just that it is one of the tokken that i missed during the initial.

I find the Nick Szabo notion of smart contract deeply flawed because , you know ... Law is Between, Code is Within. I'll post on this in the next few weeks. In a nutshell any contract dumb or smart requires ... automata parties. If it is self-enforcing, i.e. code it is all within and is not a contract.