My experience with lawyers is that if you need a legal document, they often have templates ready to go and will take 10 seconds to find & replace your company name into the document. I know this because they often don’t proofread it afterwards. They then charge you $1000 for that 10 seconds of work. Given that I have felt ripped off by many lawyers, I’m totally fine with that industry getting disrupted.
I recently had a similar experience where it became very clear that the lawyers were using AI to draft the agreements. Though it was very accurate and pedantic on some matters, it got other things completely wrong because it was not provided with full context, and full context is very difficult to provide.
The lawyers did not properly review the agreements, causing post-signature discovery of major issues and having to redo and execute all the agreements a second time, leading to inflated lawyer bills.
If the lawyer is going to prompt the same AI that I can prompt myself and is not going to review the agreements thoroughly, then it is clear where this is leading.
They surely would if they could, but since they will inevitably use it themselves they would need to collude and make a deal with the AI companies to find some way to keep the public out and law firms in (one example being targeted knee-capping of the free and low price tiers with respect to legal matters).
The ridiculous overcharging thing was something I thought the entire public was well versed in, so the idea that someone could earnestly think that they would get a discount seems absurd to me. As I recall, even the Oscars gift bags given to nominees in one of the years contained a prenup template supposedly crafted by best-in-business (divorce?) lawyers where you just needed to fill in the blanks. If you weren't getting a discount before AI, you are not getting a discount with AI I would presume.
It is understandable to see the moat as the illiteracy of the public regarding the legal system because the concerted effort to keep that illiteracy in place is a disgrace. But when the stakes are high, the moat still is finding a lawyer who knows their shit and one you can hold accountable. Which is why I find the idea of lawyers being dependent on AI much more terrifying.
A plumber comes over to repair a broken water heater. He looks at it, taps a pipe with his wrench, and it's fixed. He hands the customer a bill for $500. The customer becomes angry because all the plumber did was tap a pipe, and demands to see an itemized bill.
They can fill random document as well, it’s their expertise which template to use that fills your scenario and additionally once it gets to subtle nuances, you don’t want to do those yourself either.
The difference is that lawyers typically charge by the hour. Software is not sold based on how long it took to create.
So, if a lawyer bills $1000 for "10 seconds of work", it smells fraudulent.
If $1000 was a flat rate, then it's similar to software: you're purchasing some IP and passing off legal liability to the lawyer. How long the lawyer spent on it doesn't matter.
In the lawyer's case one might argue that it's just a time quantization inaccuracy. There's no smaller unit of time than an hour for them.
One can argue that there should be, but billing in 10 second increments might also be a bit much to ask.
One would hope that they would at least aggregate multiple smaller-than-resolution time blocks into one before billing though. (Five documents over the course of a week ends up costing the same)
The fee agreement was an hourly rate for producing a legal document. The resulting document looks like it took 10 seconds to produce. The lawyer billed $1000 worth of hours.
So, either the document was much more complex than anticipated (in which case the lawyer should have said so), or the lawyer inflated the hours.
And that would be fine, if you were getting an hour of their time. If the lawyer spent 10 minutes finding the right template, 10 seconds filling it in, and then 20 minutes proofreading it, then nobody would be complaining.
In reality they'll spend 1 minute grabbing a close-enough template, 10 seconds filling it in, 3 minutes billing you, and 55 minutes double-dipping by doing work for other clients.
I think it is more akin to paying a software contractor to make something to meet your specific need, and they give you an off the shelf product (which they may or may not have made themselves).
I think this is mostly a thing in the west. I was very surprised when I moved to Asia, and ended up emailing with lawyers back and forth for weeks, getting the right documents; the right advice, and then charged money the moment there actually was "work" to do, and I needed to hire them. It was shocking to me, because by that time I would have already spend $1000+ on a lawyer in Europe or the US.
The lawyer I ended up using for a lot of my structuring happily admitted as much when I was shopping around for someone to do what I needed. I said something along the lines of "all the other quotes I've got are obviously more expensive, I have to think that's because they've put a lot more work into them to cover/mitigate any precedent I need to be concerned with and/or are better equipped defend these documents if we end up in court". His reply was: "Oh sweet summer child! You think any of us are showing up in court to defend these contracts!? No. If someone has issues with any of these, you'll send them to me and I'll charge you my rate to deal with them. If we need use a clause in these contracts to protect ourselves or get what we're owed then I'll send a bunch of stern emails about it, and I'll charge you my rate. But if anything ever ends up in court... hah... that's not us man. Now you're having to find a lawyer that specialises in litigation. Nobody that you're speaking to about these documents right now is going to show up in court to defend these documents. That's not how it works".
When I pushed the others it turns out he was right (except for one especially large firm who basically did everything, but it may as well have been a collection of a dozen different companies). Really appreciated his honesty but was all the more baffled by how the whole industry hadn't already been disrupted.
> I know this because they often don’t proofread it afterwards.
It’s hilarious to see people complaining about lawyers being overpaid and incompetent on a forum dedicated to an industry that was completely unregulated up until about a month ago. A lawyer can be disbarred from his profession for making these kinds of mistakes; can you?
The people and their experience is still very valuable, but having been in the business since early 2000s the job shouldn’t be called ‘software developer’ anymore. Even ‘software engineer’ is somewhat inadequate given what the end product is vs what is actually being engineered by humans. ‘Software factory process automation engineer’ maybe counts?
All UK conveyancing solicitors are in this picture too. 90% of what they do is shuffle standard paperwork and chase other solicitors for their standard paperwork, all incredibly slowly. I would be entirely unsurprised if in a typical transaction, N clients' £1000*N of solicitors could be replaced by £10 of AI and a couple of hours of human time in total.
> My experience with lawyers is that if you need a legal document, they often have templates ready to go and will take 10 seconds to find & replace your company name into the document. I know this because they often don’t proofread it afterwards. They then charge you $1000 for that 10 seconds of work. Given that I have felt ripped off by many lawyers, I’m totally fine with that industry getting disrupted.
You almost certainly work in technology/software. If you (or your employers) sells software in any form, you're doing the same fucking thing.
Their box of templates is equivalent to a library or a SaaS application.
with the exception that in software there typically isn't an expectation that the client does their own debugging before accepting the software, and any bugs that manifest in said software after purchase treated as the client's fault for not spotting them.
> with the exception that in software there typically isn't an expectation that the client does their own debugging before accepting the software,
There totally is. I don't think there's any industry where a customer can really get away with not testing the products of their suppliers. With software you need to test it and confirm you're getting the results you expect, rather than blindly trusting the vendor.
> and any bugs that manifest in said software after purchase treated as the client's fault for not spotting them.
I'm not sure about B2B licenses (I've never been involved in contracts), but I'm pretty sure for a lot of B2C software, the user is responsible for the output and the vendor's liability is severely capped (like no more than refunding the purchase price).
If there is an Excel bug that costs you a lot of money, you the user are responsible. Microsoft's not going to pay you to make you whole (and I'm pretty confident about that, because logically, if it were otherwise the risk to them would be extreme and they'd probably be bankrupt).
no, it’s worse. software users will randomly be flagged to be testers of new code without their knowledge or consent. they don’t have to accept changes in SaaS, they get what you give them and they pay you for the privilege of having their software change in undesirable ways and having to contact support when it breaks.
> Given that I have felt ripped off by many lawyers, I’m totally fine with that industry getting disrupted.
I wonder how easily lawyers will let their industry get disrupted. I think lawyers may be one of the few professions that have the means and the edge to push back.
Lawyer jokes are on point and on brand for a reason.
They grift about as hard as tech workers did pre-2022. Relying on the fact that no one can easily audit their work to justify their one hour of actually work despite making easily 300K+ a year.
Please kill their golden goose. Most lawyers are actively harmful to society.
I see no evidence that in our nation's 250 year history or the broader practice of law (specifically adversarial law), that lawyers are working to make their job more efficient, legal outcomes more reliable, or legal tools more accessible.
Burn their entire industry to the ground. I welcome making the practice of law an AI-only field and ban humans from working in it to remove prejudice, misconduct, expense. As a former foster parent, I've seen children being delayed justice for weeks because two lawyers couldn't find time for a 15 minute conversation in front a judge. I've hated all lawyers ever since. Gatekeeping priesthood of the 21st century.
The priesthood metaphor works quite well as well since everything they touch ends up in a fashion that may as well have been written in Latin. And it's all wholly unnecessary since, contrary to the common perception of law operating on esoteric interpretations of a comma, lots of things in law come down to reasonable discretion. Of course exercising that reasonable discretion oneself requires first decrypting the text, which is outside the bounds of most laymen. So lawyers get paid lots of money to decipher texts that lawyers were paid lots of money to encipher using a legal system composed by politicians who were themselves, overwhelmingly, lawyers as well.
The legal profession is hardly a priesthood. If you're referring to contracts, most of the language is pretty simple. You have to learn a bit of jargon but much less than in software development. I think a lot of people are just intimidated or lazy and don't even try.
It's a critical law currently working its way through congress. It's not only indecipherable, but far longer than the entire original US Constitution. And that's a trend - when things were done by people and for people, they tended to be brief and clear. As lawyers got involved, everything became a convoluted indecipherable mess which may as well be written in Latin.
> It's not only indecipherable, but far longer than the entire original US Constitution.
I took a look at a few random sections of it and didn't see anything undecipherable, other than maybe things that lack enough context because it is written as a diff from the current law.
The comparison to the length of the Constitution makes little sense to me. The Constitution establishes a framework and some power allocations. It says almost nothing about the bazillion lower level details necessary to actual have a functioning county.
The point I'm making about the discretion is that in the past the law would lay out a general framework, and that framework would be interpreted with some reasonable discretion assumed for its legal interpretation. And that is still the case. The only difference is that now laws are written in such a way that they're largely incomprehensible to non-lawyers, and I suspect even to most lawyers outside their domain of expertise.
Obviously individual sections of a law are going to literally 'decipherable' in the sense that they are written in English, but try to tell me what that bill "really" does, the incentive for its passing, the implications, and so forth. Almost nobody would be able to do that. By contrast read the Constitution and anybody could easily understand the structure of the United States government. Or similarly the Bill of Rights lays out some of the most critical rights and restrictions in the US, yet each is no longer than a few sentences which are easily understood.
> [...] but try to tell me what that bill "really" does, the incentive for its passing, the implications, and so forth. Almost nobody would be able to do that. By contrast read the Constitution and anybody could easily understand the structure of the United States government.
OK, but now in addition to understanding the structure of US government from reading the Constitution try to understand those things you ask of the bill: the incentive for passing the Constitution, its implications, and so forth.
At the least you also need to read the Federalist Papers, the Anti-Federalist papers, many early debates in Congress over what the heck the Constitution means, many early Supreme Court cases, and probably a fair bit of history to understand the issues the country had under the Articles of Confederation that the Constitution was supposed to address.
Same for the Bill of Rights.
In modern terms things like the Constitution and the Bill of Rights are prompts. Bills and case law are implementations.
The implementations are complicated because (1) law concerns people and what they do and people and their endeavors are complicated, (2) we have a whole lot of people in a large interconnected and interdependent economy, and (3) a whole lot of people look for any ambiguity they can find and try to exploit it for their own gain, often to great detriment to others, so bills need to be explicit to cover the edge cases.
Look at the language in that bill, or any, and you'll see that when it all comes down to it, it no in way whatsoever, covers all edge cases. The law constantly uses broad strokes and relies on reasonable interpretations to clarify ambiguities. And this is the same for all bills. The law isn't like code, and doesn't try to be like code. The obfuscation is completely unnecessary.
Similarly when I speak of the motivation for things, I'm not speaking of some subtle and nuanced historical understanding of motivation, but simply what something does and why somebody might want it. If somebody is of average intelligence they can easily infer such from the Constitution, Bill of Rights, and so forth. But in that bill I referenced, you can't even ELI5 the law, let alone why or what it would change. It's an obfuscated mess. And none of this is necessary at all.
The law works because most people won't break it, most disputes don't even end up in anyone getting sued, and most lawsuits end up in settlements before too much court time was spent. Break any of those things, and systems crack. Break multiple, and it all collapses.
If the cost of what a lawyer does today drops to pennies, the system probably needs robo-judges, and/or massive penalties for taking something in front of a human judge when the facts sufficiently completely one sided, as making them pay the winning party's legal fees will become insufficient deterrence.
That is self inflicted by having a system where you're expected to plead not guilty when you're guilty, because otherwise you get screwed. In other country's legal systems, pleading not guilty when ultimately found guilty is taken as a lack of remorse and tends to result in harsh sentencing. By contrast if somebody pleads guilty, shows remorse, and works with the police and others then they can expect to receive relatively lenient sentencing, all other things being equal.
I don't think it will be a huge problem. People aren't going to commit more crimes so the caseload there won't increase much. "Jailhouse lawyers" might file more appeals of their own convictions but those seldom go very far.
On the civil side, high-dollar disputes are already litigated. The real potential growth enabled by cheaper legal services will be in small claims. Those courts can be scaled up because they don't require a jury or even a real judge.
> They grift about as hard as tech workers did pre-2022. Relying on the fact that no one can easily audit their work to justify their one hour of actually work despite making easily 300K+ a year.
>> They then charge you $1000 for that 10 seconds of work. Given that I have felt ripped off by many lawyers, I’m totally fine with that industry getting disrupted.
Ahh, the salaryman lament against contract workers (plumbers, car mechanics, lawyers etc): he makes $25/hour, why don't they charge similar or less?
Because unlike you, who get a steady stream of money even if only sitting on your ass, these people only make a living if someone calls them and asks a job done. If lawyers would charge per effective time spent as much as you would deem negligible on your side, they'd have to work a construction worker job to make a living and do lawyering on the side./
You pay the human because you assume they take responsibility in this being correct. If they literally copy-paste without checking they are no better than AI. In fact they may be worse.
Yes, lawyers have ethical and legal responsibility for their work. Plus, I'd prefer that my money not go towards the consolidation and monopolization of all intellectual labor in the hands of Sam Altman.
I think this concern is not really reasonable given that there are a vast array of LLMs, now including completely open weight frontier tier models. The valuations of certain companies is nothing but a bubble. What logic exists for it is probably in one part praying for a government granted monopoly, and in part relying on the LLM companies being deemed too big to fail when the bubble bursts. The latter part is also probably true, and a yet another good reason that too big to fail is the dumbest logic ever - rather than protecting the economy, it directly drives self destructive behavior.
The business plan of any "disruption" is to undercut the legacy competitors, drive them out of business, then jack up the prices. See also: Uber, streaming services.
In principle, that's fine, but "disruptive" businesses aren't building a positive long-term model. Rather, they are leveraging investors' money to recklessly demolish established systems and replace them with something worse.
Im a bit Darwinistic on this, like if a system can be disrupted easily, it means there are existing weaknesses that need to be addressed (taxis vs Uber, hotels vs Airbnb).
If the established system can adapt to the disruption, then as consumers we have more choices.
If the established system dies and the disruptor replaced it (DVD rental vs Netflix), then the disruptor, despite their weaknesses, do answer the consumer needs better.
If both the established and disruptive system dies, then maybe they're not that important in the first place.
I feel this is the case for a lot of industries (and even a lot of productivity-enhancers, beyond AI).
I remember a much slower change where software slowly ate away the work of property managers. Especially convenient was the ability for the tenant to perform the 6-monthly inspection remotely by uploading a half-dozen photos per room, taking just 15 minutes or so rather than waiting around for several hours for a physical meet up with the property manager. Software also allowed rent payments to be easily automated, collated and monitored.
But I don't think the number of property managers decreased, nor the price that landlords paid for their services.
When I was looking for an apartment recently, the big "luxury" buildings use AI agents to schedule tours and answer questions. Tours are also self-serve, you don't actually get to meet and talk to anyone. It's a crappy experience. I had to call an AI agent to figure out how to get in the front door of a building and it completely failed to answer this very basic question (fortunately someone eventually saw me knocking). When I toured one large complex there were several groups wandering through the building at once. It did not give me a particularly safe feeling to see that random people going everywhere in the buildings unescorted had been completely normalized.
Spoke with a banker at a well known investment bank in NYC last week. They told the law firm they want the fees cut in half or they walk. This is to a top 5 law firm. Lawyers are trying to charge the same prices, when both sides know they are working half the hours. This is even at the most premium top end, where an engagement bill can run up to 30 million.
Your epistemic barometer is way off if you think claiming "lawyers are now doing 50% less work" is the same as saying "people on the internet make shit up all the time" lolol.
Because bankers famously aren't greedy pigs, and they definitely lowered prices for their customers after the efficiencies gained with computerization, right?
The issue is that communications with A.I. are not privileged like client-attorney. There was some judge that made that ruling.
Feels like a conflict of interests to me, I'd guess friends, neighbors, children and nieces of that judge enjoying their livelihood by charging for legal services. While legal services are now available under substantially lower rates and with a lot fewer conflicts of interests from A.I. providers...
This is a completely straightforward application of existing law on privilege. To maintain privilege, among other things, communications must be confidential.
Chatting with Claude breaks confidentiality: chats with Claude are subject to arbitrary inspection by Anthropic employees, not to mention the issue of model training.
You can use self-hosted LLMs without breaking privilege. And funny enough, law firms like Latham & Watkins are now buying Nvidia GPU clusters for this purpose.
Email and phone calls are considered confidential, even though it is possible for vendors to inspect the communication. I expect AI use will be tested legally, but once precedence has been established in court it will become a common practice.
> Email and phone calls are considered confidential, even though it is possible for vendors to inspect the communication.
It's a weak analogy. Ordinary comms infrastructure providers actually have pretty robust policies, technical, and contractual measures in place that restrict employee access to customer communications. In contrast, in the wild west of generative AI, companies actively monitor session data for the content itself, in order to exploit it for their own business purposes. There is zero expectation of privacy.
So I don't share your expectation that precedent will uphold the use of consumer-tier services (in their current form) for handling privileged material.
That's Latham & Watkins, but I haven't heard of any others.
I'm not convinced that for anyone smaller than them it is worth spending on HW. For adversarial works, you will always need the latest frontier model and those things are not available for most of the firms even for licensing, and it's much cheaper to use Amazon Bedrock or similar (which is already available).
It's just not economical with the current trends. You won't be breaking any privilege even if the law firm is using a public cloud service. Maybe the law firm will have to use geofenced datacenters etc. based on deontology rules or national requirements...
But in many jurisdictions, you have to use a law firm to keep LPP, that's the point.
Don't care, saved $20,000. Superintelligence can have the details, I am too boring to threaten skynet / unplug any of it's GPUs (personal case). I'm sure it has something better to do than come after me.
But, if you were to represent yourself, are your journal entries about said case privileged? Can the opposition call into evidence a binder of legal strategies you created that morning before court?
At what point does it become a 5th amendment issue? Maybe when we get AI in our corneas a la Black Mirror's The Entire History of You?
Outside of court proceedings, a significant proportion of what law offices do is already extremely streamlined. Stuff like wills, deeds, incorporation paperwork, is all template-based and any remaining work is done by paralegals, then billed at outrageous rates. So I'm not actually sure if AI makes it a whole lot cheaper. Maybe it makes it harder to pretend.
But there's also a good reason why lawyers can keep charging these fees: because they have a say on who can practice the profession. Unlike software engineers and like doctors, they're in charge and will not allow autonomous systems to represent you or otherwise "practice law". That'd be suicide.
Also, for many services, downward price pressure already exists to some extent, in that there are various "low-stakes" virtual legal firms licensed in multiple states that provide forms you can fill and then file for $50 a pop instead of $500 a pop. But most consumers still prefer to visit a local law office. I'm not sure this will change with AI. I know quite a few people who are keen to self-diagnose using Google or ChatGPT, but I don't know as many inclined to vibecode their articles of incorporation.
It's really one of these white-collar professions that I think is quite secure. Software engineers are hosed to a greater extent, having enthusiastically engineered their own personal hell.
It's even worse with some physicians. When I got an MRI a year or two ago, getting an AI diagnosis cost extra. I am very positive about AI, but I can understand why people would be hesitant to pay extra when they hear about false positives and the unreliability of AI.
> New analysis from the Blue Cross Blue Shield Association (BCBSA) finds that hospital systems are increasingly billing patient hospital stays as more medically complex, driving nearly $1 billion in additional healthcare costs in just two years. [...]
> More than 60% of hospital systems are now using AI-enabled technologies that can scan lab results and electronic records to identify secondary diagnoses, which could move a patient into a higher-severity, higher-reimbursement billing category.
Upcoding is a tale as old as time (and one of UHG's main tactics for making truckloads of money), but this does seem new and interesting.
Like, obviously only a doctor can diagnose. But, they can use technology to help them look at problems, including AI (literally the most versatile pattern matching software we have, just in natural language and with lower success rates). And they can make their own judgement call on if what the tool output is saying is correct.
However, if a Doctor were to see a bogus secondary diagnosis on their screen, and then dismiss it, they're opening themself up to medical malpractice liability, no? Like, it seems like the safest option would be to run (and bill for) the extra tests to confirm if the AI's story is true or untrue.
From your comment, I suspect you know much more about this topic than me.
But it doesn't seem from the link that they are running extra tests, just that they are charging for the secondary diagnosis. That's odd, if there's liability for not testing or treating it, isn't it?
That…doesn’t sound like a bad thing for patients? That sounds like an increase in the standard of care. I’m not surprised a health insurer is mad about it though
Adding false secondary diagnoses is one of the most prevalent forms of healthcare upcoding and makes hospitals and health providers a boatload of money.
It sounds like AI is making doing that a lot easier, and probably hard to contest what it finds without running extra tests.
Always ask for your imaging data. I have never been denied it. Then take that and run it through your own analysis. Years ago I had some lab work done looking for a specific genetic marker that could explain something. While that was running I took my 23andme genome and cross referenced it with snpedia and had my answer (didn't have the marker) before my doctor did. She was blown away. This was 10 years ago. I'm guessing things have improved quite a bit.
Are there even any FDA labeled AI diagnostics for MRI? Just about everything I've seen is more things that are meant to help radiologists do things faster (and it's rather debatable whether they ultimately achieve that). The things I've seen gaining traction are very specific measurements (that are not themselves diagnostic) or things that improve workflow.
Setting aside the unreliability of AI, healthcare is already expensive and confusing. It's hard to know what's covered or not ahead of time and it sucks to feel nickel and dimed on something that is already so expensive, especially when you consider we are not trained medical professionals, so we can't really have an informed opinion on what this is really worth...
Legal work in the past involved hiring specialist firms, at exorbitant rates, for specialist work. (And virtually all legal work is specialized.)
Legal work in the future might involve hiring one or two in-house lawyers to handle all work, at a standard salary rate. The crucial skills are simply editing and research. Current/near-future LLMs surpass virtually all specialists in domain knowledge and they do it quickly and for the cost of a few cheap subscriptions.
You only really need a lawyer because a business can't represent itself in court without one.
My partner spent many years in big law and now sits across the table on the client side. The billable hour has been problematic for much longer than AI has been in the picture, but what many people forget (including sometimes the companies paying the bill) is that you often aren't really paying for the hours, you're paying the structural tax that has been established by the system. Hours imply work product, but sometimes the reason firms are paying outside counsel _at all_ isn't because their in-house team couldn't do it, it's because the market/courts have told them they need the liability transference granted by a signed piece of paper.
That said, the rest of the article is on point. AI billing analysis software + internal knowledge of how the sausage gets made has shaved literal millions off bills at this point. I wouldn't expect them to lose money in an absolute sense though. More like a balloon you squeeze from one end.
One thing I came here to say, but turn out not to be the case, was the large jump in associates salaries as an interesting possible driver. I found a handy tracking number that does inflation adjustment. Turns out when you index for inflation, the associate pay scales are more or less flat.
I know almost nothing about law firms. But for some other industry if you have 1000 competitors making similar products then companies will compete on price. If you reduce your prices then you will probably get more customers.
If you want to buy a new PC would you rather just have IBM making PCs and setting the price to whatever they want or a thousand different clone companies all competing on price, features, support, etc?
So some new law firm may want to disrupt the industry with a new model
Hired a lawyer. They then delegated all the work to their intern. Their intern did a terrible job. Their title in their email literally said "Intern." They didnt even hide it. The intern's work was mostly stuff claude would have chewed through in seconds at higher quality. Then they prematurely cancelled by submitting a withdrawal with the court taking my retainer. I simply built my own lawyer for my next situation.
Can’t recall where I read it, but there was an interesting post about how AI was being received by the legal industry. There were largely two camps, those who got it and knew they must adopt and adapt, and those who got it but were like “why should I adopt something that directly reduces my billable hours?” It seemed pretty clear that the latter camp was not going to exist much longer. I guess this is playing out as expected.
Think about the adversaries using LLMs to generate ever-increasing preponderous walls of text that are mostly sycophantic and subtly legally blonde. If every adversary to my law firm generates 250% more text than before, our billable hours will balloon accordingly to the time it takes junior partners to read it all, digest it, redact, whatever. In fact, speaking of redacting, perhaps billable hours increase as you proofread your own LLM output. Regardless, LLMs should empower even criminal defendants to become quasi-eloquent and "type" 1,000wpm. It would seem that "reduces billable hours" is a naïve and premature prediction.
I mean AI was being sold to law firms as a tool to automate the drudge work that takes the most time and hence contributes most to the billable hours. That is what they were reacting to. I’m not sure I follow the adversarial model here?
So… if AI is reducing time taken for billable actions, then that reduces the cost no? Unless it’s not actually reducing the time spent on said actions.
It simply put the onus of proof on the taxpayer for historical discrepancies, where the proof often no longer existed, for a subset of the population least able to navigate bureaucracy when those discrepancies were innocuous.
There was no intelligence involved at any level of assessment.
Doesn’t really sound like AI to me, more like subtracting two numbers from a government database and invoicing the difference. Was there anything done that would be classified as AI?
I think that question is very similar to why, if AI is making software developers so much more productive, those developers are not getting paid substantially more, or having to work fewer hours.
Subscriptions do seem to continue to go _up_, too. The proof of the productivity pudding is lower prices, which means the proof of all this AI hype is not so much Millennium Prizes being solved than subscriptions actually getting cheaper. If AI is already having a huge effect, inflation should be going down... if AI will have the effects OpenAI et al are currently selling, prices should plumet. I mean, I'm waiting...
Not the same thing because you are talking about employees, not customers. The way a corporation sees it is that it means they need less developers for the same (or more) delivery. And then it should mean a market more in their favor, so actually lower wages
Because not sharing the benefits of increased productivity with your employees discourages them from adopting new technologies.
If I'm 25 then sure, probably I need to learn the new technology anyway because my old technology will be obsolete before the end of my career.
Still, if I don't like your new technology, I can just switch industries and you lose an employee. You want to force all employees to learn Python while I'm okay with keeping everything in Java? Oh well, I'll find a Java job then, good luck training someone to replace me.
If I'm close to retirement, then unless you pay up significantly, there's zero reason for me to learn the new technology, and I'll resist as long as I can.
Before we got an infinite amount of AI-generated Indians, software wages were sky high exactly for that reason - in order to incentivize people to learn new technologies (coding).
Of course when it's the employer's market none of that applies and you just suck it up. But again, because minimum wage exists, if I'm working for minimum wage anyway, I'll simply choose the least demanding job where I don't have to learn shit.
One thing you get from legal council is they'll be ready to defend and contracts or agreements they've made for you. Might be hard to find defense or offense on AI slop contracts. We've seen how AI does code, imagine that inconsistency in legal docs.
But the ones drafting the documents aren't the ones fighting for the contract in court, the one who drafted slopping is happily on to the next unsuspecting client, while you bear the cost of fighting with and paying for a brand new lawyer in the courts.
There aren't any good processes that are accessible to sue lawyers for malpractice like there are for doctors and other professions, their negligence can cause entire companies to go down, but the perpetrators almost never get to pay the price.
I'm sure many law companies have a trillion spare to invest in data centres. I don't think anyone would be upset if AI broke the standard charges for lawyer fees.
My experience with lawyers is that if you need a legal document, they often have templates ready to go and will take 10 seconds to find & replace your company name into the document. I know this because they often don’t proofread it afterwards. They then charge you $1000 for that 10 seconds of work. Given that I have felt ripped off by many lawyers, I’m totally fine with that industry getting disrupted.
I recently had a similar experience where it became very clear that the lawyers were using AI to draft the agreements. Though it was very accurate and pedantic on some matters, it got other things completely wrong because it was not provided with full context, and full context is very difficult to provide.
The lawyers did not properly review the agreements, causing post-signature discovery of major issues and having to redo and execute all the agreements a second time, leading to inflated lawyer bills.
If the lawyer is going to prompt the same AI that I can prompt myself and is not going to review the agreements thoroughly, then it is clear where this is leading.
That happens even when AI isn't involved.
> then it is clear where this is leading
Lawyers lobbying to outlaw AI-generated legal documents?
*outlaw AI-generated legal documents by non-lawyers.
They surely would if they could, but since they will inevitably use it themselves they would need to collude and make a deal with the AI companies to find some way to keep the public out and law firms in (one example being targeted knee-capping of the free and low price tiers with respect to legal matters).
The ridiculous overcharging thing was something I thought the entire public was well versed in, so the idea that someone could earnestly think that they would get a discount seems absurd to me. As I recall, even the Oscars gift bags given to nominees in one of the years contained a prenup template supposedly crafted by best-in-business (divorce?) lawyers where you just needed to fill in the blanks. If you weren't getting a discount before AI, you are not getting a discount with AI I would presume.
It is understandable to see the moat as the illiteracy of the public regarding the legal system because the concerted effort to keep that illiteracy in place is a disgrace. But when the stakes are high, the moat still is finding a lawyer who knows their shit and one you can hold accountable. Which is why I find the idea of lawyers being dependent on AI much more terrifying.
This reminds me of the parable about the plumber.
A plumber comes over to repair a broken water heater. He looks at it, taps a pipe with his wrench, and it's fixed. He hands the customer a bill for $500. The customer becomes angry because all the plumber did was tap a pipe, and demands to see an itemized bill.
The plumber dutifully provides the bill:
* Pipe tapping: $5
* Knowing where to tap: $495
The other version of this joke:
Customer: $500 for 10 minutes of work? That's more than I make as a brain surgeon.
Plumber: It's more than I made as a brain surgeon too.
Except in this analogy the plumber didn't know where to tap, or knew and didn't care enough to tap in the right place.
I can tap a pipe in a random spot for a lot less than $500.
They can fill random document as well, it’s their expertise which template to use that fills your scenario and additionally once it gets to subtle nuances, you don’t want to do those yourself either.
alternatively:
pipe tapping: $5
knowing where to tap: $95
taking on the legal liability and accountability in case the water heater explodes: $400
ChatGPT is more than capable of telling you where to tap, but you can't sue openAI if it gets it wrong and your water heater explodes.
Do you have any issues with thr tech industry selling software licenses?
Because a lawyer selling a contract template, is in many ways, very similar in concept to tech industry selling a software license/seat.
They are both akin to “templates”, and selling IP.
The difference is that lawyers typically charge by the hour. Software is not sold based on how long it took to create.
So, if a lawyer bills $1000 for "10 seconds of work", it smells fraudulent.
If $1000 was a flat rate, then it's similar to software: you're purchasing some IP and passing off legal liability to the lawyer. How long the lawyer spent on it doesn't matter.
In the lawyer's case one might argue that it's just a time quantization inaccuracy. There's no smaller unit of time than an hour for them.
One can argue that there should be, but billing in 10 second increments might also be a bit much to ask.
One would hope that they would at least aggregate multiple smaller-than-resolution time blocks into one before billing though. (Five documents over the course of a week ends up costing the same)
Why would it be fraudulent if the fee agreement disclosed it?
I understood the OP's point as follows:
The fee agreement was an hourly rate for producing a legal document. The resulting document looks like it took 10 seconds to produce. The lawyer billed $1000 worth of hours.
So, either the document was much more complex than anticipated (in which case the lawyer should have said so), or the lawyer inflated the hours.
> The fee agreement was an hourly rate for producing a legal document.
We don’t know that for certain.
They bill with hourly granularity
And that would be fine, if you were getting an hour of their time. If the lawyer spent 10 minutes finding the right template, 10 seconds filling it in, and then 20 minutes proofreading it, then nobody would be complaining.
In reality they'll spend 1 minute grabbing a close-enough template, 10 seconds filling it in, 3 minutes billing you, and 55 minutes double-dipping by doing work for other clients.
Isn't it usually .1 hour increments?
They still need to proofread the doc and make sure it applies to their client though?
I think it is more akin to paying a software contractor to make something to meet your specific need, and they give you an off the shelf product (which they may or may not have made themselves).
One is transparent and the other feels like a lie.
Not in principle, but in practice.
I think this is mostly a thing in the west. I was very surprised when I moved to Asia, and ended up emailing with lawyers back and forth for weeks, getting the right documents; the right advice, and then charged money the moment there actually was "work" to do, and I needed to hire them. It was shocking to me, because by that time I would have already spend $1000+ on a lawyer in Europe or the US.
The lawyer I ended up using for a lot of my structuring happily admitted as much when I was shopping around for someone to do what I needed. I said something along the lines of "all the other quotes I've got are obviously more expensive, I have to think that's because they've put a lot more work into them to cover/mitigate any precedent I need to be concerned with and/or are better equipped defend these documents if we end up in court". His reply was: "Oh sweet summer child! You think any of us are showing up in court to defend these contracts!? No. If someone has issues with any of these, you'll send them to me and I'll charge you my rate to deal with them. If we need use a clause in these contracts to protect ourselves or get what we're owed then I'll send a bunch of stern emails about it, and I'll charge you my rate. But if anything ever ends up in court... hah... that's not us man. Now you're having to find a lawyer that specialises in litigation. Nobody that you're speaking to about these documents right now is going to show up in court to defend these documents. That's not how it works".
When I pushed the others it turns out he was right (except for one especially large firm who basically did everything, but it may as well have been a collection of a dozen different companies). Really appreciated his honesty but was all the more baffled by how the whole industry hadn't already been disrupted.
> I know this because they often don’t proofread it afterwards.
It’s hilarious to see people complaining about lawyers being overpaid and incompetent on a forum dedicated to an industry that was completely unregulated up until about a month ago. A lawyer can be disbarred from his profession for making these kinds of mistakes; can you?
did i miss something? are developers regulated since last month?
No, they are irrelevant.
The people and their experience is still very valuable, but having been in the business since early 2000s the job shouldn’t be called ‘software developer’ anymore. Even ‘software engineer’ is somewhat inadequate given what the end product is vs what is actually being engineered by humans. ‘Software factory process automation engineer’ maybe counts?
pretty sure it’s more Wild West than ever out here. engineers aren’t even reading the code anymore.
All UK conveyancing solicitors are in this picture too. 90% of what they do is shuffle standard paperwork and chase other solicitors for their standard paperwork, all incredibly slowly. I would be entirely unsurprised if in a typical transaction, N clients' £1000*N of solicitors could be replaced by £10 of AI and a couple of hours of human time in total.
I like that metaphor, because it translates well into software:
- You have a problem.
- Someone takes the time to fix that problem for $10k worth of coding time.
- Now they sell 1000s of clients the same software for $99, with 0s of work.
What's is the difference with your case?
> My experience with lawyers is that if you need a legal document, they often have templates ready to go and will take 10 seconds to find & replace your company name into the document. I know this because they often don’t proofread it afterwards. They then charge you $1000 for that 10 seconds of work. Given that I have felt ripped off by many lawyers, I’m totally fine with that industry getting disrupted.
You almost certainly work in technology/software. If you (or your employers) sells software in any form, you're doing the same fucking thing.
Their box of templates is equivalent to a library or a SaaS application.
with the exception that in software there typically isn't an expectation that the client does their own debugging before accepting the software, and any bugs that manifest in said software after purchase treated as the client's fault for not spotting them.
so yeah, not the best analogy.
> with the exception that in software there typically isn't an expectation that the client does their own debugging before accepting the software,
There totally is. I don't think there's any industry where a customer can really get away with not testing the products of their suppliers. With software you need to test it and confirm you're getting the results you expect, rather than blindly trusting the vendor.
> and any bugs that manifest in said software after purchase treated as the client's fault for not spotting them.
I'm not sure about B2B licenses (I've never been involved in contracts), but I'm pretty sure for a lot of B2C software, the user is responsible for the output and the vendor's liability is severely capped (like no more than refunding the purchase price).
If there is an Excel bug that costs you a lot of money, you the user are responsible. Microsoft's not going to pay you to make you whole (and I'm pretty confident about that, because logically, if it were otherwise the risk to them would be extreme and they'd probably be bankrupt).
no, it’s worse. software users will randomly be flagged to be testers of new code without their knowledge or consent. they don’t have to accept changes in SaaS, they get what you give them and they pay you for the privilege of having their software change in undesirable ways and having to contact support when it breaks.
Depends on the type of lawyer / work being done.
If it’s just copy paste templates then you never really needed a lawyer in the first place. Thats paralegal at best
Now AI will proofread it and quality improves.
Not really fair comparison. I agree they are ripping us, but the "template" is like a source code. It takes time and money to write and maintain it.
> Given that I have felt ripped off by many lawyers, I’m totally fine with that industry getting disrupted.
I wonder how easily lawyers will let their industry get disrupted. I think lawyers may be one of the few professions that have the means and the edge to push back.
Weird take on a site with most of us coasting on a near zero marginal cost product.
You develop software for your second client in 10 seconds?
Each additional copy? Only if it's really, really big. Otherwise it's under 2 seconds.
Your SaaS web app takes more than 10 seconds to render in your client's web browser?
Lawyer jokes are on point and on brand for a reason.
They grift about as hard as tech workers did pre-2022. Relying on the fact that no one can easily audit their work to justify their one hour of actually work despite making easily 300K+ a year.
Please kill their golden goose. Most lawyers are actively harmful to society.
I see no evidence that in our nation's 250 year history or the broader practice of law (specifically adversarial law), that lawyers are working to make their job more efficient, legal outcomes more reliable, or legal tools more accessible.
Burn their entire industry to the ground. I welcome making the practice of law an AI-only field and ban humans from working in it to remove prejudice, misconduct, expense. As a former foster parent, I've seen children being delayed justice for weeks because two lawyers couldn't find time for a 15 minute conversation in front a judge. I've hated all lawyers ever since. Gatekeeping priesthood of the 21st century.
The priesthood metaphor works quite well as well since everything they touch ends up in a fashion that may as well have been written in Latin. And it's all wholly unnecessary since, contrary to the common perception of law operating on esoteric interpretations of a comma, lots of things in law come down to reasonable discretion. Of course exercising that reasonable discretion oneself requires first decrypting the text, which is outside the bounds of most laymen. So lawyers get paid lots of money to decipher texts that lawyers were paid lots of money to encipher using a legal system composed by politicians who were themselves, overwhelmingly, lawyers as well.
Dante's 8th Circle indeed.
The legal profession is hardly a priesthood. If you're referring to contracts, most of the language is pretty simple. You have to learn a bit of jargon but much less than in software development. I think a lot of people are just intimidated or lazy and don't even try.
Feel free to have a go at this: https://www.congress.gov/bill/119th-congress/house-bill/6028...
It's a critical law currently working its way through congress. It's not only indecipherable, but far longer than the entire original US Constitution. And that's a trend - when things were done by people and for people, they tended to be brief and clear. As lawyers got involved, everything became a convoluted indecipherable mess which may as well be written in Latin.
> It's not only indecipherable, but far longer than the entire original US Constitution.
I took a look at a few random sections of it and didn't see anything undecipherable, other than maybe things that lack enough context because it is written as a diff from the current law.
The comparison to the length of the Constitution makes little sense to me. The Constitution establishes a framework and some power allocations. It says almost nothing about the bazillion lower level details necessary to actual have a functioning county.
The point I'm making about the discretion is that in the past the law would lay out a general framework, and that framework would be interpreted with some reasonable discretion assumed for its legal interpretation. And that is still the case. The only difference is that now laws are written in such a way that they're largely incomprehensible to non-lawyers, and I suspect even to most lawyers outside their domain of expertise.
Obviously individual sections of a law are going to literally 'decipherable' in the sense that they are written in English, but try to tell me what that bill "really" does, the incentive for its passing, the implications, and so forth. Almost nobody would be able to do that. By contrast read the Constitution and anybody could easily understand the structure of the United States government. Or similarly the Bill of Rights lays out some of the most critical rights and restrictions in the US, yet each is no longer than a few sentences which are easily understood.
> [...] but try to tell me what that bill "really" does, the incentive for its passing, the implications, and so forth. Almost nobody would be able to do that. By contrast read the Constitution and anybody could easily understand the structure of the United States government.
OK, but now in addition to understanding the structure of US government from reading the Constitution try to understand those things you ask of the bill: the incentive for passing the Constitution, its implications, and so forth.
At the least you also need to read the Federalist Papers, the Anti-Federalist papers, many early debates in Congress over what the heck the Constitution means, many early Supreme Court cases, and probably a fair bit of history to understand the issues the country had under the Articles of Confederation that the Constitution was supposed to address.
Same for the Bill of Rights.
In modern terms things like the Constitution and the Bill of Rights are prompts. Bills and case law are implementations.
The implementations are complicated because (1) law concerns people and what they do and people and their endeavors are complicated, (2) we have a whole lot of people in a large interconnected and interdependent economy, and (3) a whole lot of people look for any ambiguity they can find and try to exploit it for their own gain, often to great detriment to others, so bills need to be explicit to cover the edge cases.
Look at the language in that bill, or any, and you'll see that when it all comes down to it, it no in way whatsoever, covers all edge cases. The law constantly uses broad strokes and relies on reasonable interpretations to clarify ambiguities. And this is the same for all bills. The law isn't like code, and doesn't try to be like code. The obfuscation is completely unnecessary.
Similarly when I speak of the motivation for things, I'm not speaking of some subtle and nuanced historical understanding of motivation, but simply what something does and why somebody might want it. If somebody is of average intelligence they can easily infer such from the Constitution, Bill of Rights, and so forth. But in that bill I referenced, you can't even ELI5 the law, let alone why or what it would change. It's an obfuscated mess. And none of this is necessary at all.
That law is essentially a diff. The USC (the actual statute) tends to be easier to read through.
On the other hand, I'm horrified at the thought of zero marginal cost legal work given the strain of the current workload on courts.
The law works because most people won't break it, most disputes don't even end up in anyone getting sued, and most lawsuits end up in settlements before too much court time was spent. Break any of those things, and systems crack. Break multiple, and it all collapses.
If the cost of what a lawyer does today drops to pennies, the system probably needs robo-judges, and/or massive penalties for taking something in front of a human judge when the facts sufficiently completely one sided, as making them pay the winning party's legal fees will become insufficient deterrence.
That is self inflicted by having a system where you're expected to plead not guilty when you're guilty, because otherwise you get screwed. In other country's legal systems, pleading not guilty when ultimately found guilty is taken as a lack of remorse and tends to result in harsh sentencing. By contrast if somebody pleads guilty, shows remorse, and works with the police and others then they can expect to receive relatively lenient sentencing, all other things being equal.
I don't think it will be a huge problem. People aren't going to commit more crimes so the caseload there won't increase much. "Jailhouse lawyers" might file more appeals of their own convictions but those seldom go very far.
On the civil side, high-dollar disputes are already litigated. The real potential growth enabled by cheaper legal services will be in small claims. Those courts can be scaled up because they don't require a jury or even a real judge.
> strain of the current workload on courts
Maybe courts could be replaced by AI too once it's good enough?
With the right to escalate to human judges of course.
and without turning it into current phone helplines or support chatbots, god please.
How often are you satisfied with the answers from the customer service chatbot vs immediately looking for the talk to a human button?
Never. Added that prayer
> They grift about as hard as tech workers did pre-2022. Relying on the fact that no one can easily audit their work to justify their one hour of actually work despite making easily 300K+ a year.
This isn’t true at all.
>> They then charge you $1000 for that 10 seconds of work. Given that I have felt ripped off by many lawyers, I’m totally fine with that industry getting disrupted.
Ahh, the salaryman lament against contract workers (plumbers, car mechanics, lawyers etc): he makes $25/hour, why don't they charge similar or less?
Because unlike you, who get a steady stream of money even if only sitting on your ass, these people only make a living if someone calls them and asks a job done. If lawyers would charge per effective time spent as much as you would deem negligible on your side, they'd have to work a construction worker job to make a living and do lawyering on the side./
> I’m totally fine with that industry getting disrupted.
Because you'd rather pay your $1000 to OpenAI for 5 milliseconds of work, instead of to a human lawyer for 10 seconds of work?
You pay the human because you assume they take responsibility in this being correct. If they literally copy-paste without checking they are no better than AI. In fact they may be worse.
If they literally copy-paste without checking and it turns out the AI was wrong and you suffer harm from that you can sue the lawyer for malpractice.
My comment was not intended to express support of using AI instead of lawyers, but rather to express the pointlessness of "disrupting" human lawyers.
If the human lawyer doesn’t proofread it, whats the point’s the point of a human lawyer? Scapegoat for the AI?
Yes, lawyers have ethical and legal responsibility for their work. Plus, I'd prefer that my money not go towards the consolidation and monopolization of all intellectual labor in the hands of Sam Altman.
I think this concern is not really reasonable given that there are a vast array of LLMs, now including completely open weight frontier tier models. The valuations of certain companies is nothing but a bubble. What logic exists for it is probably in one part praying for a government granted monopoly, and in part relying on the LLM companies being deemed too big to fail when the bubble bursts. The latter part is also probably true, and a yet another good reason that too big to fail is the dumbest logic ever - rather than protecting the economy, it directly drives self destructive behavior.
Except you probably pay cents in tokens for that, or included with your sub.
The business plan of any "disruption" is to undercut the legacy competitors, drive them out of business, then jack up the prices. See also: Uber, streaming services.
Enjoy your cheap lawyer tokens while you can.
I mean that's just how life is? Exploit opportunities when they appear, find new ones when they run out.
In principle, that's fine, but "disruptive" businesses aren't building a positive long-term model. Rather, they are leveraging investors' money to recklessly demolish established systems and replace them with something worse.
Im a bit Darwinistic on this, like if a system can be disrupted easily, it means there are existing weaknesses that need to be addressed (taxis vs Uber, hotels vs Airbnb).
If the established system can adapt to the disruption, then as consumers we have more choices.
If the established system dies and the disruptor replaced it (DVD rental vs Netflix), then the disruptor, despite their weaknesses, do answer the consumer needs better.
If both the established and disruptive system dies, then maybe they're not that important in the first place.
Any system can be disrupted if you throw unlimited VC money at it.
I feel this is the case for a lot of industries (and even a lot of productivity-enhancers, beyond AI).
I remember a much slower change where software slowly ate away the work of property managers. Especially convenient was the ability for the tenant to perform the 6-monthly inspection remotely by uploading a half-dozen photos per room, taking just 15 minutes or so rather than waiting around for several hours for a physical meet up with the property manager. Software also allowed rent payments to be easily automated, collated and monitored.
But I don't think the number of property managers decreased, nor the price that landlords paid for their services.
When I was looking for an apartment recently, the big "luxury" buildings use AI agents to schedule tours and answer questions. Tours are also self-serve, you don't actually get to meet and talk to anyone. It's a crappy experience. I had to call an AI agent to figure out how to get in the front door of a building and it completely failed to answer this very basic question (fortunately someone eventually saw me knocking). When I toured one large complex there were several groups wandering through the building at once. It did not give me a particularly safe feeling to see that random people going everywhere in the buildings unescorted had been completely normalized.
Spoke with a banker at a well known investment bank in NYC last week. They told the law firm they want the fees cut in half or they walk. This is to a top 5 law firm. Lawyers are trying to charge the same prices, when both sides know they are working half the hours. This is even at the most premium top end, where an engagement bill can run up to 30 million.
> Lawyers are trying to charge the same prices, when both sides know they are working half the hours.
An absurd claim presented without evidence.
80% of hn comments
An absurd claim presented without evidence.
Your epistemic barometer is way off if you think claiming "lawyers are now doing 50% less work" is the same as saying "people on the internet make shit up all the time" lolol.
Which is, surprisingly, a fantastic result, considering literally all other websites of the internet.
Ngl I'm guilty myself too though.
source: my lawyer friend now orders two martinis for lunch instead of one
you want me to post their name?
Aren't bankers charging the same prices they were _before computerization_?
the value from bankers comes from relationships, sales, and their weight in doing a good IPO. its not about hours worked
Are we not still paying to “build out” the payment networks for Visa and Mastercard?
Because bankers famously aren't greedy pigs, and they definitely lowered prices for their customers after the efficiencies gained with computerization, right?
Even if you had evidence to substantiate this, a well known investment bank was never utilizing a top 5 law firm because of the rates it charged.
The issue is that communications with A.I. are not privileged like client-attorney. There was some judge that made that ruling.
Feels like a conflict of interests to me, I'd guess friends, neighbors, children and nieces of that judge enjoying their livelihood by charging for legal services. While legal services are now available under substantially lower rates and with a lot fewer conflicts of interests from A.I. providers...
This is a completely straightforward application of existing law on privilege. To maintain privilege, among other things, communications must be confidential.
Chatting with Claude breaks confidentiality: chats with Claude are subject to arbitrary inspection by Anthropic employees, not to mention the issue of model training.
You can use self-hosted LLMs without breaking privilege. And funny enough, law firms like Latham & Watkins are now buying Nvidia GPU clusters for this purpose.
> communications must be confidential.
Email and phone calls are considered confidential, even though it is possible for vendors to inspect the communication. I expect AI use will be tested legally, but once precedence has been established in court it will become a common practice.
> Email and phone calls are considered confidential, even though it is possible for vendors to inspect the communication.
It's a weak analogy. Ordinary comms infrastructure providers actually have pretty robust policies, technical, and contractual measures in place that restrict employee access to customer communications. In contrast, in the wild west of generative AI, companies actively monitor session data for the content itself, in order to exploit it for their own business purposes. There is zero expectation of privacy.
So I don't share your expectation that precedent will uphold the use of consumer-tier services (in their current form) for handling privileged material.
That's Latham & Watkins, but I haven't heard of any others. I'm not convinced that for anyone smaller than them it is worth spending on HW. For adversarial works, you will always need the latest frontier model and those things are not available for most of the firms even for licensing, and it's much cheaper to use Amazon Bedrock or similar (which is already available). It's just not economical with the current trends. You won't be breaking any privilege even if the law firm is using a public cloud service. Maybe the law firm will have to use geofenced datacenters etc. based on deontology rules or national requirements... But in many jurisdictions, you have to use a law firm to keep LPP, that's the point.
Don't care, saved $20,000. Superintelligence can have the details, I am too boring to threaten skynet / unplug any of it's GPUs (personal case). I'm sure it has something better to do than come after me.
People can’t know for sure that they saved anything until all statute of limitations have expired without a lawsuit!
They ruled that it's like a journal, or a tool like a search engine. This seems fair.
But, if you were to represent yourself, are your journal entries about said case privileged? Can the opposition call into evidence a binder of legal strategies you created that morning before court?
At what point does it become a 5th amendment issue? Maybe when we get AI in our corneas a la Black Mirror's The Entire History of You?
They'd generally be protected if you were using them purely for legal theories. But that's really left up to judicial discretion.
It's more likely to be ruled in your favor if you started the research after being arraigned, and choosing to represent yourself.
But if you were just googling "murder defense theory" and are accused of murder, they'll be able to use that against you.
Outside of court proceedings, a significant proportion of what law offices do is already extremely streamlined. Stuff like wills, deeds, incorporation paperwork, is all template-based and any remaining work is done by paralegals, then billed at outrageous rates. So I'm not actually sure if AI makes it a whole lot cheaper. Maybe it makes it harder to pretend.
But there's also a good reason why lawyers can keep charging these fees: because they have a say on who can practice the profession. Unlike software engineers and like doctors, they're in charge and will not allow autonomous systems to represent you or otherwise "practice law". That'd be suicide.
Also, for many services, downward price pressure already exists to some extent, in that there are various "low-stakes" virtual legal firms licensed in multiple states that provide forms you can fill and then file for $50 a pop instead of $500 a pop. But most consumers still prefer to visit a local law office. I'm not sure this will change with AI. I know quite a few people who are keen to self-diagnose using Google or ChatGPT, but I don't know as many inclined to vibecode their articles of incorporation.
It's really one of these white-collar professions that I think is quite secure. Software engineers are hosed to a greater extent, having enthusiastically engineered their own personal hell.
> That'd be suicide.
And yet many software developers are hyping themselves into obsolescence by pretending you can't write code by hand anymore.
It's even worse with some physicians. When I got an MRI a year or two ago, getting an AI diagnosis cost extra. I am very positive about AI, but I can understand why people would be hesitant to pay extra when they hear about false positives and the unreliability of AI.
Additionally, BCBS reports that hospitals are using AI to find ways to charge more money: https://www.bcbs.com/about-us/association-news/bcbsa-analysi...
> New analysis from the Blue Cross Blue Shield Association (BCBSA) finds that hospital systems are increasingly billing patient hospital stays as more medically complex, driving nearly $1 billion in additional healthcare costs in just two years. [...]
> More than 60% of hospital systems are now using AI-enabled technologies that can scan lab results and electronic records to identify secondary diagnoses, which could move a patient into a higher-severity, higher-reimbursement billing category.
Upcoding is a tale as old as time (and one of UHG's main tactics for making truckloads of money), but this does seem new and interesting.
Like, obviously only a doctor can diagnose. But, they can use technology to help them look at problems, including AI (literally the most versatile pattern matching software we have, just in natural language and with lower success rates). And they can make their own judgement call on if what the tool output is saying is correct.
However, if a Doctor were to see a bogus secondary diagnosis on their screen, and then dismiss it, they're opening themself up to medical malpractice liability, no? Like, it seems like the safest option would be to run (and bill for) the extra tests to confirm if the AI's story is true or untrue.
From your comment, I suspect you know much more about this topic than me.
But it doesn't seem from the link that they are running extra tests, just that they are charging for the secondary diagnosis. That's odd, if there's liability for not testing or treating it, isn't it?
That…doesn’t sound like a bad thing for patients? That sounds like an increase in the standard of care. I’m not surprised a health insurer is mad about it though
Adding false secondary diagnoses is one of the most prevalent forms of healthcare upcoding and makes hospitals and health providers a boatload of money.
It sounds like AI is making doing that a lot easier, and probably hard to contest what it finds without running extra tests.
UHG owns the whole apparatus. They are very happy with cost increases.
I didn’t quote it, but they also say at the link it didn’t change treatment.
That’s why they think it’s BS
Always ask for your imaging data. I have never been denied it. Then take that and run it through your own analysis. Years ago I had some lab work done looking for a specific genetic marker that could explain something. While that was running I took my 23andme genome and cross referenced it with snpedia and had my answer (didn't have the marker) before my doctor did. She was blown away. This was 10 years ago. I'm guessing things have improved quite a bit.
Are there even any FDA labeled AI diagnostics for MRI? Just about everything I've seen is more things that are meant to help radiologists do things faster (and it's rather debatable whether they ultimately achieve that). The things I've seen gaining traction are very specific measurements (that are not themselves diagnostic) or things that improve workflow.
Yep one imaging center in LA is charging $45 out of pocket for the AI diagnosis.
Setting aside the unreliability of AI, healthcare is already expensive and confusing. It's hard to know what's covered or not ahead of time and it sucks to feel nickel and dimed on something that is already so expensive, especially when you consider we are not trained medical professionals, so we can't really have an informed opinion on what this is really worth...
can you say no to an AI diagnosis in that case?
Yes, sorry should have mentioned that. Of course the easy solution is then to simply decline.
There's a solution that most are overlooking.
Legal work in the past involved hiring specialist firms, at exorbitant rates, for specialist work. (And virtually all legal work is specialized.)
Legal work in the future might involve hiring one or two in-house lawyers to handle all work, at a standard salary rate. The crucial skills are simply editing and research. Current/near-future LLMs surpass virtually all specialists in domain knowledge and they do it quickly and for the cost of a few cheap subscriptions.
You only really need a lawyer because a business can't represent itself in court without one.
My partner spent many years in big law and now sits across the table on the client side. The billable hour has been problematic for much longer than AI has been in the picture, but what many people forget (including sometimes the companies paying the bill) is that you often aren't really paying for the hours, you're paying the structural tax that has been established by the system. Hours imply work product, but sometimes the reason firms are paying outside counsel _at all_ isn't because their in-house team couldn't do it, it's because the market/courts have told them they need the liability transference granted by a signed piece of paper.
That said, the rest of the article is on point. AI billing analysis software + internal knowledge of how the sausage gets made has shaved literal millions off bills at this point. I wouldn't expect them to lose money in an absolute sense though. More like a balloon you squeeze from one end.
One thing I came here to say, but turn out not to be the case, was the large jump in associates salaries as an interesting possible driver. I found a handy tracking number that does inflation adjustment. Turns out when you index for inflation, the associate pay scales are more or less flat.
I know almost nothing about law firms. But for some other industry if you have 1000 competitors making similar products then companies will compete on price. If you reduce your prices then you will probably get more customers.
If you want to buy a new PC would you rather just have IBM making PCs and setting the price to whatever they want or a thousand different clone companies all competing on price, features, support, etc?
So some new law firm may want to disrupt the industry with a new model
https://archive.is/dmUuD
Law firms are going to be rolled over by AI. The inefficient law firms will be replaced. Like Expedia, etc. did to travel agents.
Hired a lawyer. They then delegated all the work to their intern. Their intern did a terrible job. Their title in their email literally said "Intern." They didnt even hide it. The intern's work was mostly stuff claude would have chewed through in seconds at higher quality. Then they prematurely cancelled by submitting a withdrawal with the court taking my retainer. I simply built my own lawyer for my next situation.
Can’t recall where I read it, but there was an interesting post about how AI was being received by the legal industry. There were largely two camps, those who got it and knew they must adopt and adapt, and those who got it but were like “why should I adopt something that directly reduces my billable hours?” It seemed pretty clear that the latter camp was not going to exist much longer. I guess this is playing out as expected.
> directly reduces my billable hours
Think about the adversaries using LLMs to generate ever-increasing preponderous walls of text that are mostly sycophantic and subtly legally blonde. If every adversary to my law firm generates 250% more text than before, our billable hours will balloon accordingly to the time it takes junior partners to read it all, digest it, redact, whatever. In fact, speaking of redacting, perhaps billable hours increase as you proofread your own LLM output. Regardless, LLMs should empower even criminal defendants to become quasi-eloquent and "type" 1,000wpm. It would seem that "reduces billable hours" is a naïve and premature prediction.
I mean AI was being sold to law firms as a tool to automate the drudge work that takes the most time and hence contributes most to the billable hours. That is what they were reacting to. I’m not sure I follow the adversarial model here?
So… if AI is reducing time taken for billable actions, then that reduces the cost no? Unless it’s not actually reducing the time spent on said actions.
Reduces the cost to the medical provider. You think that'll result in a reduced the cost on the person's bill?
For a lawyer yes, if a task takes less time they bill less hours and therefore it costs less.
I’m just waiting for the AI judgments, followed by the flash trials, followed by the flash litigation DDOS attacks.
Australia already has AI tax judgements:
https://en.wikipedia.org/wiki/Robodebt_scheme
Hard to call it "AI".
It simply put the onus of proof on the taxpayer for historical discrepancies, where the proof often no longer existed, for a subset of the population least able to navigate bureaucracy when those discrepancies were innocuous.
There was no intelligence involved at any level of assessment.
Doesn’t really sound like AI to me, more like subtracting two numbers from a government database and invoicing the difference. Was there anything done that would be classified as AI?
https://www.youtube.com/watch?v=XL2RLTmqG4w
“Cyst O’Donavan has been found guilty. It took a Waymo jury less than three seconds to convict him.”
I think that question is very similar to why, if AI is making software developers so much more productive, those developers are not getting paid substantially more, or having to work fewer hours.
Subscriptions do seem to continue to go _up_, too. The proof of the productivity pudding is lower prices, which means the proof of all this AI hype is not so much Millennium Prizes being solved than subscriptions actually getting cheaper. If AI is already having a huge effect, inflation should be going down... if AI will have the effects OpenAI et al are currently selling, prices should plumet. I mean, I'm waiting...
Not the same thing because you are talking about employees, not customers. The way a corporation sees it is that it means they need less developers for the same (or more) delivery. And then it should mean a market more in their favor, so actually lower wages
why would developers be paid more if it's the capital investment (i.e. the machines) that's leading to more productivity?
Because not sharing the benefits of increased productivity with your employees discourages them from adopting new technologies.
If I'm 25 then sure, probably I need to learn the new technology anyway because my old technology will be obsolete before the end of my career.
Still, if I don't like your new technology, I can just switch industries and you lose an employee. You want to force all employees to learn Python while I'm okay with keeping everything in Java? Oh well, I'll find a Java job then, good luck training someone to replace me.
If I'm close to retirement, then unless you pay up significantly, there's zero reason for me to learn the new technology, and I'll resist as long as I can.
Before we got an infinite amount of AI-generated Indians, software wages were sky high exactly for that reason - in order to incentivize people to learn new technologies (coding).
Of course when it's the employer's market none of that applies and you just suck it up. But again, because minimum wage exists, if I'm working for minimum wage anyway, I'll simply choose the least demanding job where I don't have to learn shit.
Sounds like a case of "Where Are the Customers' Yachts?" again. As brought up by an author called Fred Schwed regarding Wall Street investment.
AI is just a tool like everything else. Are you also going to ask me for a discount because I used a powered screw driver?
Yes, because I'm expecting you to compete.
Exactly, somebody using power tools can do a lot more in an hour than somebody using manual tools. That is the real discount.
since we're posting pointless gotchas:
Do we pay more to a lawyer if they handwrite a legal document using expert penmanship?
and conversely, I'll pay you more if you grab a rake and not a leaf blower
yes, because you're billing me based on hours.
One thing you get from legal council is they'll be ready to defend and contracts or agreements they've made for you. Might be hard to find defense or offense on AI slop contracts. We've seen how AI does code, imagine that inconsistency in legal docs.
But the ones drafting the documents aren't the ones fighting for the contract in court, the one who drafted slopping is happily on to the next unsuspecting client, while you bear the cost of fighting with and paying for a brand new lawyer in the courts.
There aren't any good processes that are accessible to sue lawyers for malpractice like there are for doctors and other professions, their negligence can cause entire companies to go down, but the perpetrators almost never get to pay the price.
The whole law should have been automated and formally verified decades ago.
That's basically like devs being 10x with AI - they ain't getting a raise for that..
it's the AI that's 10x not the devs, why should the devs be getting a raise?
Lawyers and politicians are the first things that ought to be replaced by AI, honestly.
Remember that it can always get worse.
We've watched Terminator and The Matrix. We're ready
Well we’ve spent over a trillion on infrastructure and we’re only just getting started so good luck getting your fucking discount
Don’t really get the downvotes here - if you spend more on AI than you ever did on payroll your clients aren’t going to get a discount.
I'm sure many law companies have a trillion spare to invest in data centres. I don't think anyone would be upset if AI broke the standard charges for lawyer fees.