Breaking the law

(in to pieces)

Robots are not replacing lawyers, nor politicians. People, maybe, will be able to do it.

Many of us fear that humanity will end with metal creatures dressed like judges dictating final verdicts to suspects that have not even committed crimes. But, there is a long run until we get to this sweeping scenario. Many choices could be made so we can keep control of technology and it help us live in society. There are many ways for improving the control that we have over our democracy. Voting is a tiny piece of this huge chain. We can go further, specially by maintaining and versioning our own rules of operation. (the rules of our system)

Our laws, I mean.

We are in the very beginning of participatory processes at the law making ecosystem. As collaborative environments to work with text are transforming the way we share our thoughts and code, laws, and their repositories still tend to be closed and unreachable for most of us. Anyone can follow free software documentation, install and use — which means to undergo specific rules of usage and behavior. It is an agreement. But, with laws things are different. Unfortunately, democracy, with all the potential to be a real agreement, is just one more instrument of oppression and control.

The good part is that now we have the internet, design and code. Laws, bills and other instruments of bureaucracy urge to be updated to this beautiful environment so we can create, change and deprecate laws whenever we need to.

Some, at this point, may ask:

“OMG, if we could create our own laws, the world would be a mess. People would make laws approving indiscriminate use of guns, drugs and other harmful things. And the bad guys would allow others to kill and rape, and steal from those who have nothing. People would make laws to allow slavery!”

Well, behind some beautiful dresses, this bad things are already happening everyday, with or without government regulation or control. I’m sorry to say that in real world, outside democracyland (whatever this may be) these terrible things are unstoppable and unreachable by politics, police or even state surveillance. I could explain here why, but I’ll just suggest the use of newspapers and Google Translate. Brazil, at least, is a country that routinely violates human rights and no one cares because everybody is working and producing good commodities. So let’s get back to laws.

In the wild, there is a lot of studies and new platforms for voting, debating and even to suggest laws and rules. I’d place more emphasis on those who are free and easy to install, like Wordpress. Sometimes small solutions are effective and handy on having significant public participation. But more sophisticated solutions, like DemocracyOS — that aims to generate better arguments for our rulings between other things — can also bring transparency and structure to the entire flow of lawmaking. Advantages of this little additional gift: adherence, giving sustainability to the process.

Besides outnumbered platforms for collecting inputs in the law-making process, something terrible happens when there are “sufficient” amendments to the initial bill. All this suggestions goes to a void, that is actually similar to a liquefier: someone uses the input to build another version of the proposed text, no one knows exactly why the choices were made.

This means that the necessary merges are personal choices.

In other words, there is a lack of transparency maintainers at the regulatory processes. No one is showing the diff nor being transparent at the choices to absorb inputs to proposed laws that were supposed to be changed at platforms for public consultation.

In addition to that, quantity also means bottlenecks: when you have a lot of comments and few people to do the rewriting work, time and quality are compromised.

Imagine that a public consultation received over a million comments: how does all this comments would be absorbed? This happened in India, when TRAI launched public consultation over Net Neutrality. Part of the concerns against the results are related to the process — manually handled, according to the authorities.

People who collaborated in public consultations actually want to see their footprints in the final deliverable. When they are not contemplated in this desire, at least giving them some explanation on WHY would be elegant — and mandatory for democracies (as we want them to be).

So the questions are:

1) how to give power over the results of public consultations for those who collaborated?

2) how to make all the process transparent?

So that laws are not baked like pizzas nothing better than the duetto “data and design” to give users (citizens) more options, all through free data and free software.

The idea could be to deliver a dashboard-like experience over enriched data, linking meaning and suggesting new approaches over text at some point. For executing this, a good initial point would be analyzing the experience of merging texts. Observing people working with google docs to go thru comments at a proposed bill, I identified some groups of actions:

elements of processing and comparing text manually

Comparison, highlighting, association, merging, mapping, connecting by meaning, and other practices, — all well known by machine learning — are easy to identify as starting point to interfaces that allow transparency at merging. Still, deep learning by itself won’t do the magic. At the Research Center on Web Technologies (Ceweb) we are conducting an exploratory research with a do and think approach. Partnered with the InWeb, a group of academic researchers with knowledge at machine learning and Web and with the Ministry of Justice, we are trying to figure out how to explore the results of a public consultation when data is enriched and explored as a public asset, allowing the citizen to interact with the results of the process. As we just initiated the project, there is still a lot of research and software to develop. So far, we offered tools to help in the manual merging of the results of the consultation, but Interface Design and experience are critical, since we are talking about using the web as a tool to do something.

There is a lot of people and organizations doing similar stuff around. Last week, when the new draft of the Privacy and Data Protection was posted by the Ministry of Justice, collectives were fast in providing simple solutions for comparison, like this diff, brought by Lucas Teixeira (from Coding Rights). He also provided a python tool to compare texts, including some singular characteristics of Portuguese as a language. Again, simple and handy tools doing their magic.

Other initiatives to benchmark are those brought by the Cap Collective. Among a series of projects related to Open Democracy, the DebatesCore aims to integrate public consultation with semantic tools. Some initiatives aims to open all the diff log between new and deprecated versions of laws and regulations. I’ve been at ISWC and saw a presentation of Michael Gus Creedon and William Bill Tanner entitled The evolution of a thesaurus-driven semantic web search service for delivering the written prose of Government Policy. This guy aim to provide an updated version of laws and regulations with semantic linking, using RDF model. They have a wiki-like tool that offers to the users the possibility to search and find semantic connections between expressions.

For discovery of diffs, the best government experience that I have found was the UK initiative on Legislation. The motivation of the changes at the laws are not clear (no commit message :-/), but the interface to search for changes is very intuitive.

Still, as I am a unconditional lover of Git and the way it works, I miss a lot of transparency and integration with the data related to all the lawmaking ecosystem. Of course, some of the best features of systems like Git are not available for those who don’t code. This is so important that Github, a Social Website that is a layer between humans and Git, is considered by some people as critical infrastructure.

Moses Destroys the Tables of the Ten Commandments

I’ll keep searching for solutions that prioritize the user as a citizen when dealing with public participation. Gathering a huge amount of suggested interventions at a proposed rule can be considered success, but only if the next step is asking again and again, until consensus is reached.

And for those who say that consensus is not compatible with democracy, I could say that at least it **should be**.