"... Think of what version control software could mean for the Congress [...]. If bills were created under a system where strike-throughs and additions were carefully tracked, the public would know which legislator made which change to a proposed piece of legislation as it made its way through the Capitol.
At last, there would be transparency in the legislative process. Best-case scenario, it would shame legislators from inserting language against the public interest and only meant to reward political contributors; at worst, it would make such insertions public and allow the voters to punish the politicians who made them. ..."
There are a million frontends designed to simplify Git for non-programmers. In fact, lots of office software now has version management built-in. Very few people use it. This isn’t a technology problem.
To be fair, version control built into office suits are often hidden away or have other UX patterns that aren’t comfortable for nontechnical folk. So they either still get scared by it, or don’t even know it exists.
Yet there is a demand for version control for nontechnical people. And I know this because I’ve seen enough files called “copy of important thing” or “important thing 2”.
> Yet there is a demand for version control for nontechnical people. And I know this because I’ve seen enough files called “copy of important thing” or “important thing 2”.
This demonstrates LACK OF demand for technical version control, since the system you describe has been used since forever and found to be good enough.
No. It demonstrates nontechnical people adapting the skills they’re familiar with to suit a requirement they have.
It doesn’t prove that they don’t want a better solution should it have the right UX. And, as I said in my original comment, the current UX for version control in documents both sucks and is often hidden away where people might not even be aware it exists.
Yes, it is, but the issue is ease of use for non-technical users.
I was trying to avoid naming any products on purpose, but, for example, there is a commercial product that is easy enough for most people and affordable that would work for this- and there are a few open source clones that come close.
The open source clones are not quite polished enough for non-technical users, but they are the closest thing I can think of, basically crossing over a wiki + Markdown for knowledgebase creation with Git-like versioning.
We don’t need version control for legislative bills. We need journalists that track down who tried to insert what, summarize it, and present it to the public with context. That would satisfy the oversight purpose without getting lost in technicalities.
Our problem is that the press isn’t doing their job, not that legislation doesn’t have a commit graph. The latter, even if it existed, would not fix the former.
A revision control system is a tool the journalists could use to achieve this. Commits to the 'enacted' branch could have commit messges documenting the context (when, how, why, who). Unmerged branches could represent proposed but ultimately rejected (or still in progress) changes with the same amount of meta data.
Of course a tool itself doesn't solve the problem. You need somebody to use the tool. But the tool makes it easier to solve the problem.
Until journalists can be paid under the table to mislead the public, the problem will persist. Almost anyone can be bought on a Friday night for the price of a few drinks and benefits.
I am confused. Someone put in place a system that would allow journalists or citizens to meaningfully engage with the law in a much better way than they can now.
However you are proposing that journalists as a homogenous group do some extra work. I think that's a great thing but I don't get why you say the press aren't doing their job when they are at an historical low of compensation and respect.
If sounds like you are outsourcing your rage at the legislature and the press and journalists are the problem?
Sticking to the FTA someone made a first step in version tracking legislative texts but you do not think they've done enough?
We need that too, but the ability to git diff changes to the law is a great win for democracy and transparency in 2026 because git is an interface that agents are experts at using. If you're a journalist, activist or anybody else doing the type of research you just described this makes it faster and easier.
In the present day, mainstream journalists won't be meaningful advocates for democracy and the public interest any time soon. It's time to acknowledge that this idea is wishful thinking. The root cause is changes to the law like the Telecom Act of 1996 which allowed for the consolidation of old media under a few corporations.
For many years these mainstream journalists have been required to support corruption, cronyism, the centralization of power, and the increase of corporate power in order to retain their jobs. The law needs to be changed before they can become a net benefit to society again.
> because git is an interface that agents are experts at using
I'd actually argue that this is the wrong way around. Being able to git diff legislation is very good, but mainly for humans, who need tools to make work tractable.
An agent will use whatever interface is available, including no interference, won't give up due to the task being too onerous, and can be scaled by paying for more tokens. It could simply brute force diff the raw text, writing bespoke Python and regexes, etc.
LLMs make brute forcing intellect possible. It's humans who need finely crafted tools.
I am sure with LLMs the time to craft those tools to fetch the data and put them back in a git repo (for example one repo per legislative project) would not be too hard to do.
> Our problem is that the press isn’t doing their job
The general public isn't willing to pay for this, so it's very hard to do this. Especially given that most of this isn't exactly going to generate tons of views.
Let's be honest, in general, the public is very upset about all the injustice they perceive. But the public also lost all interest in the process, some can't even be bothered to vote once every other year....
It's too easy to blame others for your own inaction
> But the public also lost all interest in the process, some can't even be bothered to vote once every other year....
In the 1920s the German speaking world got a book whose first three chapters just hammer in how important it is for the average citizen to take an interest in his local politics. That book was subsequently banned by almost every European country in the decades since. I'm the only person I know who has personally read it, even though it is one of the most derided books in history.
I highly recommend the first three chapters to anybody who thinks they don't need to understand their local politics.
I don't think that the problem with legislative transparency is that there aren't enough middlemen sitting between the public and the legislative process.
I'm personally not interested in being dependent on the press -- which is just other people pursuing their own interests and assumptions -- to interpret artificially opaque sources, if those sources can just be made accessible instead.
Having a dynamic and pluralistic press is a good thing, and we need more investigative journalism, to discover what isn't visible to the public, but I'm not sure there's a lot of value in the sort of interpretive journalism you're proposing.
Git does allow multiple contributors on each commit. Of course people can lie, but it's a matter of policy in the end. Make it a crime to not add all contributors to a diff.
Every system gets gamed, and so is the current one. The goal is that the people can more easily stay informed. Even with mixed authors problem you highlight would be still an improvement.
This is already regularly done. One side or the other will propose or amend a bill that they claim does [popular thing x] and then add in some clause that they know will get voted against, so they can then claim that [other side] voted against [popular thing x].
This is often especially cynical because [popular thing x] is often something that neither party wants to pass. So a very recent example of this was an act against insider trading by politicians, that was bundled with a voter id law. Neither side wants to clamp down on insider trading from politicians, so the entire act was DOA. It was nothing but bread and circuses.
The politicians wouldn’t have the option to game it if additional data is pulled in automatically like lobbying disclosure filings, public watchdog databases, etc. for the final overview. A swift turn from limiting the total sugar amount in soda would look quite bad if the politician got lobbied by Big Soda.
> it would shame legislators from inserting language against the public interest and only meant to reward political contributors
You’d just got a system of favors for who inserts what language. You’d hand tough edits to your safe-seat colleagues and insert messaging amendments designed to get voted down for own comms.
When a measure becomes a target yada yada.
(Note: this proposal is for the U.S. code. Not bills. Git for the USC works.)
Such a requirement would not do what you think it would. There would not be names legislator or even aids as commit authors. The commit authors would be members of the House Office of the Legislative Counsel[1].
The House Office of the Legislative Counsel is non-partisan, impartial, and confidential. Whether transparency has merit or not, tracking provenance would make the office partisan. Is more partisanship in government not a concern?
The second issue with transparency is that communication between legislators and legislative counsel member is subject to the attorney-client relationship. Breaking attorney-client privilege is a serious decision. Knowing how such a change would affect the organization and its members is not something to be taken lightly.
A more convincing argument for transparency would address impartiality and confidentiality. This is pretty clearly a Chesterton's fence type situation where advocating to change a system should anticipate criticisms. Addressing opposition preemptively is an implicit demonstration of knowing why the system exists which is table stakes for changing it.
There are several alternatives, like for example a political party where 1000 random citizens are picked up for analyzing and voting on every file. Next file next 1000 batch of random citizens are picked.
We could also let people vote directly on their own laws without intermediaries.
Sounds like a sort of jury system added to the legislative process. That might actually be a workable idea as an additional safeguard/sanity check before bills can pass.
2. It will create an incentive structure that will reflect the goals of political parties.
Transparency brings the office into the political sphere. Nothing in the political sphere escapes partisanship. It's pretty clear that no one knew this office existed before I brought it up and folks thought that lawmakers drafted their own bills. The fact that we're even now talking about it in terms of politics is evidence that it will devolve into partisan politics.
I get your instinct - transparency is in general a net good. But the more public an office is, the more it is drawn into partisanship.
Politics is especially prone to engineers disease and this is a great example of it. It assumes a technical solution in the spirit of transparency applies to complex human, social, and political domains where those frameworks do not neatly fit. When someone points out engineers disease, have some introspection before questioning their motives.
Sponsor of change should obviously be shown as at least partial author of the commit. Tag of "new book of laws" should show as authors all who voted yes.
Thank you for this informative comment, and I agree with the general idea in your Chesterton's Fence point.
When the HOLC provides wording, they do so in response to a request from an elected member or that member's staff, right? What I'm getting at is that it doesn't matter whether provenance is found via the commit author or via some other change-associated metadata -- the key thing is making precise change provenance trackable and (perhaps after some predefined delay) public. This would mainly aid journalists.
Is the HOLC currently functioning as an anonymizing service for changes to in-progress legislation? And if so, is that by design, or more by gradual and organic evolution?
Fundamentally, if we can't get at that provenance information in a reliable way right now at all, then version-control-style tracking would be a pretty big qualitative change in how things are done: information that wasn't available, except through unofficial channels, would now be available (again, maybe with some time delay or whatever, but in the long run it would be available). On the other hand, if that information is currently obtainable but is just not easy to obtain, then this would be more of an incremental improvement: since journalists and others with the requisite know-how could get the information right now, but maybe have to jump through hoops to do so, this would just reduce the number or height of the hoops.
I don't know which of those two scenarios best describes reality, but maybe you do?
That was the same for the software patent directive, we had leaks of amendments proposed by big companies, we could see that Members of Parliament were just copy pasting them.
By this point "the public" just doesn't care for this sort of stuff, we just want (stabler, for those still employed) jobs and day-to-day life that is not expensive as hell anymore. What the middle-class elites now call "democracy" has just become a glasperlenspiel, i.e. nothing of real interest to us, the masses, just a social differentiator between said well-to-do middle-classes.
More generally, the sooner us, the masses, spell it out that democracy is well and truly dead the better, that way maybe we can do something to alleviate our situation, but I'm afraid that won't happen anytime soon.
Bills are written in English but I'd view them as code. It's not interesting to track changes in code for outsiders. Sure, it's cool for programmers, but users of software do not care if you unrolled a for-loop and who did it. They don't even know what that is and the same holds for bills. We have journalists for that and nobody is willing to pay for that because nobody gives a shit about a functioning democracy. Now that's one hell of a problem git ain't going to solve.
The great promise of crypto-contracts was to be able to write rules as code and have them be deterministic (instead of requiring a judge to rule on it) and even have unit tests. Instead it seems we are moving towards a world where rules are written in English and judged by biased LLMs.
I couldn't care less about crypto currencies, but crypto contracts had the potential to replace much (although not all) of the bureaucracy around governments and justice systems. I am very sad it didn't take off.
There's a good chance it could have the opposite effect. The public doesn't pay as much attention to the workings of the government as donors do.
It's essentially an extension of the Sunshine Act of 1976, which made congressional votes public. There were claims that before it was enacted, legislators were able to vote their consciences, but then lie to donors and tell them they voted for their interests instead of their constituents' interests.
Since it passed, the public has been paying attention, but not like the donor class has been. Something needs to be not only bad, but have some sort of novel or salacious quality to make the news. If it's not near an election, people forget, if they paid attention to begin with. Donors pay attention. Representatives know this.
Ultimately, the problem is the money in politics, but this sort of thing can exacerbate the effects.
is money the problem or just that most people don't care about most things, but a few people REALLY care about a few things? There's plenty of groups without much money that still gum up the works because they're persistent.
People worrying about money in politics are overemphasizing the instrumentality of money, rather than the underlying structural incentives that tend toward special interests dominating complex issues.
If money were removed from the equation, the concentrations of power, motivations to influence it, and asymmetries in those motivations will all continue to exist; political influence will just be exercised through other means, and some particular faction or class of people will still be at the sweet spot of having the greatest competence and the greatest motivation to apply those means.
A lot of people are reluctant to acknowledge it, but this is inherent to politics itself, and there is no straightforward way to fix it, other than to limit the reach of political power.
It's slightly more complex though isn't? Because the actual problem is how to limit the reach of political power while still having an effective government.
In a vacuum I absolutely agree with you. In fact my PoliSci 101 prof defined politics as “who has power and how they use it” but in a system like ours where money also is a form of power (or at least a force multiplier of power) unto itself, money helps concentrate political power into fewer hands and magnify the ways they can use it.
It certainly would be naive to think that simply “getting money out of politics” would “fix” things but it is still a desirable goal for anyone who, themselves, does not currently have extremely large quantities of money.
As long as there are people who allocate billions with a sign of the pen, with zero personal responsibility and skin in the game, there will be whole industries dedicated to swaying their hand.
I don't think this is necessarily a good policy. I'm saying it's obviously blatantly untrue to suggest that just because people will always want to corrupt public officials, that they will be equally successful in all possible universes.
It's also untrue to suggest that any intervention will necessarily create a worse outcome than the status quo.
We need to amend the Constitution to strike pardon power and give the Congress the power to pass property taxes (aka wealth taxes) and regulate donations to campaigns, candidates, parties and partisan causes.
We also need law establishing mandatory retirement for Congressmen and judges, including justices, and a system where each President appoints (with Senate consent) one and only one justice to SCOTUS. Finally, yes, pass McCain-Feingold, pass a SCOTUS ethics bill and make corruption charges easier to bring.
This is an observation about the world in general, not U.S. politics specifically. Prosecuting elected officials basically means that you trust the bureaucracy more than voters. This has not proven to be a successful approach.
We have been very lucky in the U.S. to avoid these problems. And it's not because nobody before Trump ever did anything rising to the level of paying off a porn star and calling it "legal expenses." It's because we have had a taboo against prosecuting politicians.
You can't police that. The only thing you can reliably do is allocate power. If you allocate power to prosecutors to prosecute elected officials, you've basically handed them the power to decide who people are allowed to vote for. This power gets abused in virtually every country where it's normalized.
The safeguards within the justice system are appropriate for prosecuting individuals. They are not set up to handle matters of national importance that could way the direction of the country or deprive tens of millions of voters of their preferred candidate.
If you don't believe that, then consider a hypothetical system where we allow prosecutions according to the standard you propose. Except only Texas courts and prosecutors have jurisdiction.
> 1) An 18-year (9*2) term is more compatible with age limits than a 36-year term
2) Each Senate class gets to participate in the appointment of one justice
I like No. 1.
No. 2 seems overcomplicated compared with staggering retirements in the way Art. 1 § 3 cl. 2 staggers the Senate. The Senate is also more diverse than the Presidency; limiting the latter would check against some dirty tricks.
It's amazing how you've managed to position "politicians were pursuing their own hidden agendas and then lying about it" as a desirable state of affairs.
It sounds like they were defrauding people into giving them donations under false pretenses, and the transparency measures were effective at mitigating that.
There's quite a lot of strange bias in your comment: the unjustified disjunction between "constituents" and "donors", the concept of a "donor class" as some singular bloc rather than an arbitrary aggregation of anyone who donates on behalf on all sides of all issues, the idea that donations are more important than the actual votes, etc.
That seems like an assumption, and a counterintuitive one: wouldn't you expect someone getting "fleeced" by the political system to be motivated to pay attention due to that very fact?
If you have a system of open transparency, then everyone who is motivated to pay attention has the means to do so. If they're not paying attention, it means they're not sufficiently motivated, and are choosing not to pay attention, not being constrained artificially.
This seems to be converging to some sort of "false consciousness" argument, where someone speculatively projects motivations and intentions onto other parties, then criticizes those other parties for not acting in accordance with the motivations strangers have projected onto them.
Needless to say, these sorts of arguments are utterly disingenuous, and in this case, somewhat ironic, because the people arguing that outcomes should conform to their speculative ideologies -- even in defiance of what the public themselves are manifestly choosing -- are themselves special interests trying to exert their own influence over politics.
> It sounds like they were defrauding people into giving them donations under false pretenses, and the transparency measures were effective at mitigating that.
There are no pretenses in political giving (at least at the U.S. federal level), because it is illegal to tie a donation to a particular position, vote, or action by the politician. It’s called a quid pro quo and is the test for whether a donation is a bribe.
In the U.S. the political system is based on people, not positions or policy. Donations to a politician are to the person. It’s not a proxy for buying an outcome.
In practice, of course people are going to support politicians who they like and think agree with them. But under the law, a politician can’t rip you off by taking your money and then voting contrary to your desires.
Which makes the initial promise to vote a certain way in exchange for donations itself illegal. That seems fine. So what is the problem with transparency, again?
This subthread is a discussion about pertains to politicians misrepresenting their voting intentions in order to get donations. My comment was a response to an argument that allowing politicians to do this is a good thing because it allows them to vote their "consciences" while pretending otherwise to people donating to them.
Perhaps not precisely, but it is a proxy for attempting to influence an outcome.
> under the law, a politician can’t rip you off by taking your money and then voting contrary to your desires.
Isn't that in conflict with:
> it is illegal to tie a donation to a particular position, vote, or action by the politician.
Are you saying that if I give a politician money and express my desire that they are then legally bound to vote according to my desire? Sounds like buying votes; I'm pretty sure that's not condoned by the law.
In practice there is a donor class and constituent class who are largely not the same set of people. The point you’re missing is how the donor class is instrumental in influencing the actual language that goes into the laws, this happens peripheral to the democratic process. For example, a spending bill where the language favors one vendor over others because that vendor paid to submit their suggested language. This is not hypothetical btw.
I don't know what the answer to public votes is. I see the benefit in them voting their conscience. But I also see how they can vote against my interests and then lie through their teeth that they are helping me if I can't see their record.
But voting today is such a performative act for them. There are great pieces of legislation that can't pass because someone has labeled it as radical so one side or the other can't be seen with their name on it so they don't pass it even though they know it is good and would benefit their constituents.
Git commits get rebased, squashed, cherry-picked and edited retroactively, written during pair or mob programming sessions or rewritten during code review. And we also know that commit messages often suck. As a result, commits almost never tell you the full story about who exactly contributed, in what order or why.
The process is even messier when writing new bills. They get passed around between different departments, discussed & edited in various committees, etc.
Sure, you could force all participating parties to use git, prevent them from force-pushing or rewriting history even on their "feature branches", get them to sign their every commit, etc. But what you'd then likely just end up with is most negotiation & editing happening out-of-band.
Besides: At the end of the day it's the (majority of the) parliament that enacts a given law, so in my book they are all responsible, even if in practice laws get written by a small subset of people.
I'm blown away by the hand wringing of this. Someone would be forced to answer for the way they voted and inserted language. It makes it harder to do what you complain about because someone is going to take a hit. If its a little guy he might lose the next election. If its a big guy, eventually these things add up. Not sure what unintended consequences might appear. It sounds like you are resigned to the worst fate right now.
It shows git-like diffs plus an AI summary of what changed. There is no author attribution though - I'm not sure there is a business use case for it, lawyers don't care about who wrote a specific law.
The Hacker News Community (hah) needs to lobby hard to pass a law (what boostrapping!) to require the https://en.wikipedia.org/wiki/Office_of_the_Law_Revision_Cou... to maintain the repo as part of their duties, and do it retroactively for the entire history of the US Code.
It's a simple idea; it's cheap; there are no outstanding technical questions; it benefits almost everyone. Let's get this done.
It's easy and fun for volunteers to put in the effort once. It's doing the work on an ongoing basis that is less fun and less motivating. The ongoing work --- done without modern tools --- is already that office's responsibility, so this should be too.
>> The Hacker News Community (hah) needs to lobby hard to pass a law (what boostrapping!) to require the https://en.wikipedia.org/wiki/Office_of_the_Law_Revision_Cou... to maintain the repo as part of their duties, and do it retroactively for the entire history of the US Code.
> I want to see a practical benefit to maintaining this with public funds versus the usual redlines.
The US government works for the people and, as such, changes made to the US Code over time must be traceable, discoverable, and therefore justifiable by or condemnable to those who are responsible for said changes without fear or favor.
The data is "there" in some some sense, but it is messy, coarse-grained, and arrives slowly.
Once the office is familiar with Git, we can start the conversation about having the new version of the US code instantly ready when a bill is passed. In particular, rather than voting on the informal patch, congress should vote on the final result. (bills are patches, but votes should be on the new state with the patch applied.) Then the office merely supervises the Git repo, and is freed up to focus on other codification tasks, such as hyperlinking definitions, overseeing formalization of the tax code like France did, etc.
Previous discussions often foolishly conflate version control with "no force pushing of any branch" and "all branches must be public". That is stupid. Vote trading will continue to make drafting happen in secret, since it is simple game theory that an "atomic transaction" of many possibly-unrelated things is an efficacious coordination mechanism. This is just about getting the final version and its already-public history cleaned up and mechanized to the greatest degree possible.
Poland has an official, somewhat manual version of this - the ISAP system.
Actual bills are written in the form of patches ("§1. In the law X paragraph 5, the words "such and such" are replaced with "this and that". §2. The law X paragraph 4 is striked. §3. After law X paragraph 8, a new paragraph 8a is added that says "etc. and so on".)
After a patch or a collection of patches are passed, they are combined together and a unified version of the law is published. These documents aren't technically the law, they just show a snapshot of what the law is at the specific moment.
ISAP system is an online collection of all bills that were passed, plus crosslinks between laws that change/supercede/otherwise interact with each other, plus current and archival unified versions, plus related administrative orders and other miscellanious documents. Makes looking up laws and changes to laws very easy.
First PDF is the original, second PDF is the unified version, then there's links - 4 others changed, 4 explicitly obsoleted, 12 implicitly obsolete, 707 administrative orders, 12 Constitutional Tribunal rulings, 11 previous + current unified versions, 212 patches, 14 relevant EU directives.
I've been working with government entities for document management for 15 years now, writing software that does exactly what we are talking about - versioning, tracking, publishing legislative records both through the authoring and approval cycles and then to the public. I have not worked at the federal level, but with small government - school districts, small towns etc.
Those years tell me 2 things:
1) Yes, absolutely version control for legislation is critical.
2) That doesn't mean Git is the right answer.
Since I don't work at the federal level, I don't know what system they use, but I do know that their workflow is already based on specific changes - they will write up the changes to a document, then apply the changes and publish a new document. Those documents are posted on congress.gov. So they do have a versioning workflow and public accessibility, just not the same style as software devs are used to. I'm sure it could be improved, but I'm not sure forcing them into our working style is the right answer.
The real goal here is not "Use Git". That is an implementation detail, and one that has also been presented and discussed quite a bit in my niche of the industry. Mostly, it is unwanted. Or was unwanted the last time the discussion went around. Git makes sense to software engineers, but not to the general public, and not to any of the document authors I've worked with. Maybe we can add a UX layer to fix that, but that isn't the bigger point...
What we want is to work towards the actual goals: For the tracking to be more detailed, with more transparency including not only authorship of the change, but the source of the change. And presumably, some commentary as to why the change was made. Then all of that should be available to the public.
If git works as a data structure that supports a solution that meets those goals, fine. I'm not anti-git. But when designing solutions, we need to start from the goals of the system, not the data layer. And we need to design solutions that the document authors will actually use, not force them to work like us.
I hadn't thought of using git to track the changes in legislation. This is a cool idea.
In Australia, I spent a bit of time trying to work out how to find which politician voted for what (both at state and federal level), so that when elections come, you can single out the individuals who voted for what legislation. I think AI may make this easier, but the challenge is in not spamming the "free and open" access to the online record of our legislation.
So far, it's probably a one-person $50K per year job to do this task manually, and the language of the public legislation record didn't lend itself easily to determining the laws or rights that actually changed when bills were passed into legislation. They are explicitly stated, but hard to distill with regexes, filters or other analysis tools (e.g. "Who voted to stop electric scooters going over 25km/h?") If you try reading the legislation, you will know what I mean.
"A complete and open record of the proceedings of our parliament" still needs people spend a good chunk of time interpreting and summarising it (e.g. the media, commentators, etc) which means it's a time-consuming hobby or a paid gig.
> In Australia, I spent a bit of time trying to work out how to find which politician voted for what (both at state and federal level), so that when elections come, you can single out the individuals who voted for what legislation.
That's an interesting project - I am guessing it's only targeting federal level, but could also include state level
The first problem I have in my mind is proxy votes, and votes that are made mandatory by the party they belong to - eg. a backbencher would struggle to go against the wishes of their party unless they felt it was worth risking their career over
> "A complete and open record of the proceedings of our parliament" still needs people spend a good chunk of time interpreting and summarising it (e.g. the media, commentators, etc) which means it's a time-consuming hobby or a paid gig.
Same. I don't know if it's the word choices or just me recognizing it as AI and subconsciously feeling cheated.
As an aside, my city FB group(small town, basically the local forum and news) has been absolutely flooded with AI 'flyers', and I have the same reaction. I can't put my finger on the exact reason, maybe the busy-ness, but they're repelling to my eyes.
Luckily after about 6 months of this now they're starting to get heavy pushback.
I've found that AI-written text consistently fails to comprehend the context. It either assumes there's little context available, so it writes too much, or it assumes the reader has all the context, so it writes too little. Human-written text has those problems far less often.
I hope the LLM builders are paying attention and they add some kind of step in the process that causes the system to reflect on its assumptions about context and work to refine its understanding of the context before generating output.
I'm doing this as open source for all amendment-driven law for the world.
Finnish/Estonian law basically solved now but US/UK are a lot more complex to do correctly for some reason.
Now major version coming within 1 month probably with many architectural changes. The thesis is that it's basically all doable as a compiler but there may be certain cases which are irreducibly ambiguous and need human/AI to adjudicate them. The rest must be deterministic.
Cool project! I take it "LawVM" as a name and "frontend" as a term are plays on LLVM and its different frontends?
Two suggestions: 1) Handwrite the website (mainly the copy but also the design) instead of letting Claude do that. Right now it is really hard to read and comprehend. 2) Drop the "frontend" terminology. Even if it makes sense from a technical perspective, people working in the legal field (who I suppose you want to reach) will likely find it confusing.
Yeah, like LLVM. Compiler for legal text-state -> structural analogy to other compilers (just need to reverse-engineer the exact specifications at the same time).
I think I'll call it "frontend" still internally but I'll look for better terminology.
Agreed re the website. Initially had a simpler website and well, I guess I was trying to achieve something re getting actual jurisdictions interested on adopting it and gauging how much energy to spend on this in general.
Hopefully one day soon my other projects are a lesser priority so I "get to" do the websites again, and maybe figure out what I'm even trying to communicate, exactly or achieve with the website. Hopefully global adoption eventually.
I can't help but feel like this is the wrong way to interpret legislation. The Code isn't some sort of program you run to get your outcome. There's people and "intent" and societal vibes involved. Amendments end up bearing on each other, and changes to one law might change the relative weighing of it related to some other law.
It's all messy human stuff that I don't think "diff" handles all that well.
If you're like me and also eager for cool uses for git, I'll share a cool one I use -- try out git for your resume! Commits change as you progress in life, and make branches for individually tailored resumes.
A surprising amount of tracking and diffing in legal texts is actually very different to git. It's a trap to assume software version control fits the needs for legal work.
Let's say you have some case law, and to know if it's relevant you need to pull the primary law/codes and regulations .You also need to know when the correct versions of everything are, so the case law needs to resolve to the correct version of the correct piece of legislation. "What is an atomic unit of legislation?" is also fuzzy as amendments will both move and edit named blocks of legislation at once. It's a big messy very hard to track pile of moving targets, and you need to account for the temporal aspect while linking documents not only to the changes but also to documents at certain revisions. Standard git and other version control has trouble expressing many of these concepts, and while it could be hacked in, I have found it more effective to create a purpose-built tool for the job.
I'm not an expert in this domain but I'm aware there has been prior work to attempt this. I'd appreciate if someone knowledgeable could post some of that context.
Something akin to this is a big part of the products that companies like LexisNexis or Thompson Reuters charge your lawyers a whole lot of money for. Because for a lot of legal questions, you don't just need to know what the law says now, but what it said at the time of the incident in question. And every legislative body around the country (at least at the time I was working in this space a decade ago) publishes legislative changes in completely different and often times not easily parsable forms. Web scraping and PDF scraping was far more common than you'd think. Some state legislatures published changes in more reasonable forms, and some states contract their official legislative compilation to companies like LexisNexis or Thompson Reuters and work to publish an actual change stream with proper standardized references. There's a lot of manual work that goes into making stuff like this work, a lot of lawyers and other legal professionals manually reviewing that the right pieces were removed or added in the right places, matching up references to parts of the law that might have changed between the time the legislation was passed and the time the change was published and figuring out what order everything actually goes in. And the actual text of the law is only part of the equation. Legal decisions from courts can dramatically impact the reading of the law, or effect effectively strike laws from the books (even if the laws are technically still on the books, just not enforceable). So often the right way to represent a diff of two points in time depends on the question you're trying to answer.
I wanted to do something like this for a single section in Washington law, to show why a particular funding restriction had been added, but was no longer achieving the intended goal. Git alone wasn't enough of a UI, because the 'why' was only in the introductory text of the bill that changed the law, and not in the law itself - and then there was the move of that article to a different section, and a re-numbering of sections, and a change in definition in a different section that affected the scope of this one... fascinating stuff!
I thought legal document versioning software would cover some of these problems but it doesn't seem like a similar enough process.
Following renumbers has to be one of the worst parts of this. As an example, the text of the US code that was modified by the BCRA in 2002 has been renumbered / moved since Citizens United v FEC was decided. But the text itself is the same as it was since the BCRA passed, because courts don't remove the actual law text with their decisions. So if you want to know what parts of the current US code were ruled unconstitutional by the courts in that decision and are therefore unenforceable, you need to first find the text of the BCRA, because the ruling refers to the sections of that piece of legislation. Then you need to find the version of the US code that was on the books when the BCRA was passed to find the section references that the law talks about. Then you need to find the law that renumbered that part of the US Code and find the new parts. Then you need to see if any changes have been applied since (there hasn't really) and then you need to match the changes to the referenced sections in the old version of the code to the equivalent text in the current version of the code, modulo any changes to references that changed as part of the renumbering but are otherwise the same.
I'm not saying lawyers aren't really really expensive, but you don't have to spend a long time digging into just a simple question like "what parts of the law are no longer enforceable after a supreme court decision" before you start to understand why they charge the prices they do.
Cool idea but I can't find anywhere on the linked page to try it? It reads like a neat concept but it's not clear if this is a thing they've actually made or not (or am I just missing something super obvious?)
Sorry about that. There is a small-scale mock-up https://lexifina.com/watch which is what was on the video. I'll provide full public access/forkability in a few hours, watch this space.
I had this idea over 10 years ago but never finished it. I hope this takes off. I do fear it's not very practical for many big pieces of legislation which are often written in secret.
Likewise, a younger, more idealistic me thought it'd be great to use a GitHub-esque platform for laws. Pull Requests to suggest laws, gated behind reviews/approvals, branches so each house could work on their own versions before finally reconciling and merging into the Code, literally.
Transparency has a cost. There are several good reasons, but my favorite is intellectual freedom. Policy makers would be much less likely to explore bold reforms or risky bipartisan compromises if their draft quality notes and correspondences were publicly tied to their political reputation. It takes significant work to produce successful legislation, and much of that work can be productively and safely done without direct public oversight. This is the same reason an individual might value a private journal.
If you have a problem with your representative’s behavior, call their office and ask about it. They will certainly have a reason. Tell them if you disagree, and vote when you can.
> Policy makers would be much less likely to explore bold reforms or risky bipartisan compromises if their draft quality notes and correspondences were publicly tied to their political reputation.
Only if those bold reforms weren't in the interests of the people they serve. Risky bipartisan compromises should still be transparent along with what was gained in exchange for them. If the people feel that the compromise wasn't worth it they should be empowered to vote accordingly come election time. It's the job of representatives to assure the people who elected them that they are doing a good job and if they can't do that they deserve to lose their position. Refusing to support bold reforms that the people they represent want, or refusing to compromise when it would be beneficial to voters would become just as costly to their career.
> It takes significant work to produce successful legislation, and much of that work can be productively and safely done without direct public oversight.
Often most of that work is done entirely for them by corporations and lobbying groups and "law makers" just vote how they are paid to vote without even reading the bill put before them. Just because something can be done in secrecy doesn't make that the ideal way to do it. We've have extremely poor results with the system as it stands right now, so a change to introduce more transparency and accountability seems like a very good idea.
> This is the same reason an individual might value a private journal.
A public servant representing the people by making edits to laws and performing the duties they were elected to carry out is nothing at all like someone keeping a private journal. Their work is not their personal affairs, it is the public's affairs. How that work is being done matters to the public. Legislators would still remain free to go home at the end of the day and write in their private diary all they like.
> If you have a problem with your representative’s behavior, call their office and ask about it.
When their behavior is conducted in secrecy behind closed doors the public doesn't have the ability to hold them accountable for it, which is exactly why it takes place in secrecy. Increasing the public's ability to hold elected officials accountable by increasing transparency can only improve outcomes for voters, and it would certainty be inconvenient for elected officials who want to put their own interests above the interests of the people who elected them.
Unlike code, natural language laws have many exceptions and context dependencies, so trying to manage them cleanly with Git seems like it would be tough as data structuring costs and exception handling (hardcode-style responses) increase.
> natural language laws have many exceptions and context dependencies
The complexity is in common law, not natural-language laws.
The U.S. Code and Code of Federal Regulations can be completely presented through a Git-like system. The complexity comes in when attaching court cases.
I see. The laws are completely different from my country, so this is very educational. My country uses civil law, but I see that is how it is under common law. Thank you very much.
If you've got a statute like "no vehicles in the park," the fundamental issue is that you have to interpret "what is meant by 'vehicle'?" and no legal drafting is going to be good enough to actually cover all the possible cases well enough. The main difference between a common law and a civil law system here is that in a common law system, you have to cite the previous similar cases (because precedent is binding) whereas in a civil law system, you're merely using those previous similar cases as persuasive evidence. At the end of the day, it boils down to arguing for your interpretation.
And in my experience working on language standardization (which is closer to a civil law system than a common law system), it's fairly often that I have to resort to looking up not just when a change in the text occurred, but dive deep into the discussions of the paper proposing the change to understand what the committee thought it was doing with the change. The complexity of law is fairly similar--the changes to the law themselves often require understanding the legislative and interpretative history of the law to understand what the intent of the drafter was to be able to rule on how the law is extended to an unanticipated scenario.
Yeah but none of this matters for the "git for legislation" idea right? The point here is to record the history of the text of the laws, not their meaning.
It depends on the question you're trying to answer. The text of the US law which Citizen's United struck down remains on the books today (though has changed since it was moved to a different part of the code, so references have been updated), but the actual effective law has most certainly changed from when it was written to now. Since courts can't write or modify the law, the literal text remains, even if some of that text is effectively stricken as if the text had been removed.
If you want to know the literal text (that may or may not be enforceable), then sure a plain text diff is sufficient to know that nothing at all has changed about federal communication laws since this section was last amended in 2002. But if you want to know what the actual law is, you definitely need more than just a text diff.
See my response to the parent, but git blame solves the issue of "what did they mean" because the person (if alive) can just be asked. The same way git blame is used all the time to ask why a programmer did something some certain way, there's more to code too than just how exactly does it compile/execute. The higher level idea should be to stop legislation seemingly written by nobody, that nobody wants to answer for, but is used to make important decision about lives and livelihoods. Having courts needed to interpret verbiage that was written by living people is completely preposterous. If it's written by dead people, sure, that's a harder problem.
> If you've got a statute like "no vehicles in the park," the fundamental issue is that you have to interpret "what is meant by 'vehicle'?"
Maybe. Maybe it’s a court finding the legislature never had the power to pass that rule. Maybe it’s what constitutes a park.
Visualizing changes in statute with git is fine. It would have real difficulty in presenting changes in law, as a whole, as well as an annotated supplement. (At the end of the day, redlines, git trees and annotated supplements are just visualization methods. The actual federal law lives in the USC, CFR and opinions of our courts.)
> Unlike code, natural language laws have many exceptions and context dependencies,
In the interpretation of the law, yes, but that doesn't matter for version controlling the text of the law, and in that sense, you can absolutely manage the text pretty cleanly in git. Most laws are just text. Many bills are literal git-diffs, except in prose, which can make it hard to understand a bill.
> exception handling (hardcode-style responses)
This is basically non-sequitur (there no such thing as a "hardcode-style response" in git…), and again, I think you're misunderstanding the idea here.
This brings us to my even more unlikely idea - when whoever wrote that line in the legislation dies or retires, the line enters a short review period where it's either renewed with a new author on the line to speak to it or is deprecated. This way someone would always be there to just answer for the intent. Also would help garbage collect legislation that is no longer useful.
> when whoever wrote that line in the legislation dies or retires, the line enters a short review period where it's either renewed with a new author on the line to speak to it or is deprecated.
That seems like a great way to turn functional legislation into swiss cheese as one line at a time things are either kept or removed based on chance/circumstance and the whims of the current people in power.
Laws are collectively voted for as a complete package, and a law that worked and was passed as being acceptable can become useless, dangerous, and undesirable with a very small number of lines removed from them. I want lawmakers to spend their time focused on the entirety of a bill. Your idea would result in them fighting over who should or shouldn't be recorded as the official author of each and every line. We don't need nonsense like "Senator X will retire soon so Senator Y should author it!", or "I'll agree to author line 405 if you agree to author line 30!"
Your idea, like so many ideas where laws just fade away unless something happens to keep them, also sounds like a great way for you to lose longstanding freedoms and protections that should never be at risk of simply deprecating just because some one gets voted out or dies or retires.
The intent of a law should be clear from the text. It isn't always, but at least now it's possible. To have the intent of a law, or even worse the intent of specific lines within a law, subject to change at any time would make interrupting that law and establishing precedent a nightmare for the courts.
We really should have a system in place to help clear out old and unnecessary laws on the books, but this would be a terrible way to accomplish that.
Wow! This is very cool. I had an idea for something very similar to this when I found out the US federal government offered API keys, maybe three or four years ago. I found that while Congress did list bill revisions as they were made the format and retrieval of these was not an easy task. The idea was that I wanted to look at any line of any bill and see who proposed it that particular line’s inclusion.
For the record, and feel free to steal it if you like it, I called it “G(overnm)it-blame”
Cool project, good thread. Tracking activity in the US is a big challenge. A company we built helps hundreds of non-profits try to keep track of the happenings at the Federal and State level. States are actually where most of the game is played out these days. Let us know what you think of https://billtrack50.com
I love this, and I feel every legal document (of which there are many, for example, a random one: insurance companies' policies) should have something like this (and no, Word is not good for it)
YES PLEASE, but I suspect it will degenerate into another paid service with 1000 tiers... I think you cannot access the US law without paying for it? odd... I will wait for claude law to appear and hope they share the torrent...
Yes, it should absolutely change (both should be freely available), since ignorance of the law is not a defense, and case law is law (here, in the US, since we're a common law country). But the flip side of that is that the law has to be knowable, if ignorance is not a defense, or you're just going "lol too bad for you for not knowing the secret law".
Also, because how the judicial branch decides case law needs to be reviewable by the People, so that we can inform the various legislatures to adjust the law if the judicial system is not ruling to the will of the People.
Even without AI it should be. Honestly the less AI has to do with giving people access to our laws and court records the better since the value of having that access would decline with hallucinated laws/interpretations/attribution. A post about how AI isn't even fit to write wikipedia articles just fell off the front page a few hours ago, and providing something like this to the public seems at least as important.
Given that chicken scratch in the margins gets into bills that are voted upon, it seems unlikely that this would be accepted by Congress. Love it though.
I proposed Git and Markdown instead of XML for EU law making in the European Parliament, they went with their own custom web editor to manupilate XML. And failed.
The tl;dr is that doing git-style change-tracking of the legislation isn't as helpful in practice as you'd think (especially because a lot of the very basic change-tracking is already available in the existing tools that lawyers use!).
I put a lot of thought into this last decade, I find that some bills are big blocks of new text, and other bills are convoluted instructions to modify other blocks of existing law, sometimes saying "strike this word and that word" other times referencing just a subsection of another law. hyperlinks to that other law or a model of what the legal code would look like is necessary
Agentically this could be done now, but it would still just be the tip of the iceberg
You're looking at quantum states of law and regulations, and that's just for one jurisdiction
Without 64-bit time though, can't use it historically. Since current legislation is always changes built upon legislation from long before unix epoch, it kind of falls flat.
Sounds neat, but not something a private company should be responsible for.
If transparency is the goal, the entire codebase for this should be public domain and not profitable. Otherwise there's a profit motive to rewrite history.
Not really. It is arguing for version controlled legislation. So who’s responsible for what portions of literal laws and policy. I hope your "not clone-able" is in jest
That’s like saying that the code in a repo belongs to the whole team once a PR is merged, so it doesn’t make sense to track its provenance back to its original authors.
> so it doesn’t make sense to track its provenance back to its original authors
The legislative process is inherently collaborative in a way that's ordinarily undocumented. Who introduced an amendment and who supported it are entirely separate. Someone who introduced language can wind up not supporting it because of where the mix of trade-offs winds up.
At the level of a bill, knowing who wrote language matters. At the level of law, you should only care who voted for it. (I'm not even tackling the inherent subjectivity in tracing e.g. language in the CFR to specific language in a bill.)
even then you have three-line whips and numerous reasons to vote for / against things that are wrapped up in party politics and other stupid games
once something is passed; the intent of the writer, the intent of those that voted for it and the meaning of the words in the final legislation can all be wildly different.
There is complexity behind the vote, yes. But the vote is what matters. Who initially introduced text into a bill that did or didn’t pass, with or without their vote, is on the legislative side. That’s a lot more complicated to track than the USC.
There's actually two kinds of changes here: changes to the law, and changes to your description of the-law-and-its-changes (metachanges). For example, if you make a mistake in recording a change to the law, you need to fix that (unlike eg a source repo where mistakes in a commit message are forever).
So to make this a git repo, I guess you'd make the snapshots be the law plus markup for all the changes, and the diffs are the metachanges. For each snapshot S, you could generate a git repo whose snapshots are the law and the whose diffs are the changes in the law (as described by S). However, this probably isn't the best way to query S.
"... Think of what version control software could mean for the Congress [...]. If bills were created under a system where strike-throughs and additions were carefully tracked, the public would know which legislator made which change to a proposed piece of legislation as it made its way through the Capitol.
At last, there would be transparency in the legislative process. Best-case scenario, it would shame legislators from inserting language against the public interest and only meant to reward political contributors; at worst, it would make such insertions public and allow the voters to punish the politicians who made them. ..."
(from https://www.nytimes.com/2007/07/23/business/media/23link.htm... )