With the technological development, more and more processes, documents and contracts are being filled using online methods. Do you think this can replace traditional immigration lawyers?
Not entirely. There's always litigation over the areas of law that are not clearly defined, or situations that are edge cases.
In theory, a well defined system of law adhering to strict construction should allow for a systematic implementation free from bias and greatly reduced in errors. But our system is so complex and full of contradictions that I don't see that being possible.
> With the technological development, more and more processes, documents and contracts are being filled using online methods. Do you think this can replace traditional immigration lawyers?
No, not anytime in the near future.
It can perhaps reduce the need in some of the simpler cases where they are in demand, and increase the caseload each can manage effectively.
If you're just wondering about immigration lawyers, you should reflect this in the title. The answer to the universal question is "not anytime soon". The answer to the specific question might be different. As a former lawyer, I suspect there are certain aspects that can be automated, but that the subjective parts cannot be.
In general, lawyers already 'automate' some of this work by offloading it to lower-paid staff. As lawyers become more tech savvy, they will probably offload more to computers/applications than to people. The crop of senior lawyers that are about to retire have relied heavily on secretaries/paralegals in a way that junior lawyers do not. Junior and mid-level lawyers use technology much more already.
I think a lot of the rote gruntwork in the legal process has been automated already. The people that suffered from this are probably the paralegals and lower tier entry level lawyers.
Once you go beyond that, my (distinctly non-lawyer) experience has been that there is a lot of subjective territory involved.
For example, my social circle includes a lot of immigrants. I've seen cases that should have been easy get rejected because they decided to cheap out with the lawyer, or even attempt to do the filing themselves. On the other side, I've seen cases that everyone was expecting to get rejected, pass, because they afforded a good lawyer.
For drafting up a will, probably for many people. Software is pretty good at workflow management. For handling a dispute? I don’t think so. Not until apps can achieve some kind of general intelligence. There’s a lot of nuance and judgement.
Justin Kan tried it with Atrium and failed, despite having experience and top tier backing. It sounds like a hard problem, but you can stalk him and ask what went wrong.
I think it is unlikely for the reasons given below. Despite already-existing automation of legal tasks, the demand for lawyers, the work available for lawyers, and legal fees are all on the rise.
It's interesting to look back to the legal market of the early- to mid-1990s, when the market for new lawyers and lawyers in general was soft, and then the USA experienced the tobacco litigation and Microsoft antitrust litigation. This soaked up thousands of lawyers working "overtime" and saved the doomed careers of so many junior to senior lawyers who were un- or underemployed. Nowadays, with ML-based data processing and database/search tools, it's not inconceivable for a handful of lawyers to handle this work.
A smart person, whose name I don't recall, analyzed whether lawyers will be replaced by technology, and broke lawyers' roles into four categories:
1. PROCESS - Drafting contracts, preparing other paperwork, filing things. It is conceivable that a lot of this can be replaced, but (a) clients have a knack for doing things that don't fit into a mold, (b) contracts need to get revised, which is usually too messy for AI/ML, and (c) the lowest level and lowest paid attorneys are still miles ahead of AI/ML on this front, so there is not a big cost savings. This is viewed as a necessary training ground for irreplaceable (see below) senior lawyers. Robots can play soccer just as poorly as children, but if you replaced children with robots, then you would never get your next Messi.
2. ANALYSIS - Reviewing and understanding changes in codes, statutes, regulations, and case law, plus their implications for the future. You could also include writing those codes, statutes, and regulations. This function will not be replaced by AI/ML in our lifetimes.
3. ADVOCACY - Appearing before courts, tribunals, regulators and administrators, arbitrators, and other forums, and even government clerical workers to advocate for clients. This includes the judges themselves and their staff. Obviously this human element will not be replaced.
4. COUNSELING - Listening to and understanding the client’s needs, knowing the pathways, strategies, and tactics to get your client to achieving their goals, and advising them. This ranges from merely being a person to whom the client can vent all the way to the thousands of dollars per hour lawyers who devise strategies to eliminate tax and liability for mega-corporations and billionaires. Almost no human wants to tell their story to a robot. No AI/ML in our lifetime is going to be able to understand what another human’s needs are, ask the right questions and elicit the needed responses, while showing empathy. No AI/ML in our lifetime will have the trust of a CEO or senior manager.
There is a fifth function that smart person didn’t consider:
5. “INSURANCE POLICY” - Hiring a lawyer wins you the right to sue them for malpractice if they screw up. Because of this, lawyers are very thorough, cautious, and conservative, often vetting untested ideas and risk-carrying work through many layers of their law firms. Successful legal malpractice suits are rare, but the threat of them and the reputational damage makes legal malpractice itself very rare.
Could a robot replace immigration lawyers? Maybe in the PROCESS aspect of things (filling out the forms). After all, TurboTax is perhaps already as good as an accounting firm for the vast majority of Americans’ taxes, and you could perhaps say the same of Willmaker Pro compared to an estate planning law firm for most Americans. But that seems unlikely, because, unless the mechanical process of filling out a tax form, you have to interview the people, who want a human to listen to them, understand the employer’s needs, and come up with a strategy. The client wants a human to explain it to them and comfort them as to their odds. If the AI/ML screws up, only a human can fix it (low probability) at multiples of the cost of hiring them in the first place. Remember, immigration law ...
8 comments
[ 3.4 ms ] story [ 33.5 ms ] threadIn theory, a well defined system of law adhering to strict construction should allow for a systematic implementation free from bias and greatly reduced in errors. But our system is so complex and full of contradictions that I don't see that being possible.
No, not anytime in the near future.
It can perhaps reduce the need in some of the simpler cases where they are in demand, and increase the caseload each can manage effectively.
In general, lawyers already 'automate' some of this work by offloading it to lower-paid staff. As lawyers become more tech savvy, they will probably offload more to computers/applications than to people. The crop of senior lawyers that are about to retire have relied heavily on secretaries/paralegals in a way that junior lawyers do not. Junior and mid-level lawyers use technology much more already.
Once you go beyond that, my (distinctly non-lawyer) experience has been that there is a lot of subjective territory involved.
For example, my social circle includes a lot of immigrants. I've seen cases that should have been easy get rejected because they decided to cheap out with the lawyer, or even attempt to do the filing themselves. On the other side, I've seen cases that everyone was expecting to get rejected, pass, because they afforded a good lawyer.
It's interesting to look back to the legal market of the early- to mid-1990s, when the market for new lawyers and lawyers in general was soft, and then the USA experienced the tobacco litigation and Microsoft antitrust litigation. This soaked up thousands of lawyers working "overtime" and saved the doomed careers of so many junior to senior lawyers who were un- or underemployed. Nowadays, with ML-based data processing and database/search tools, it's not inconceivable for a handful of lawyers to handle this work.
A smart person, whose name I don't recall, analyzed whether lawyers will be replaced by technology, and broke lawyers' roles into four categories:
1. PROCESS - Drafting contracts, preparing other paperwork, filing things. It is conceivable that a lot of this can be replaced, but (a) clients have a knack for doing things that don't fit into a mold, (b) contracts need to get revised, which is usually too messy for AI/ML, and (c) the lowest level and lowest paid attorneys are still miles ahead of AI/ML on this front, so there is not a big cost savings. This is viewed as a necessary training ground for irreplaceable (see below) senior lawyers. Robots can play soccer just as poorly as children, but if you replaced children with robots, then you would never get your next Messi.
2. ANALYSIS - Reviewing and understanding changes in codes, statutes, regulations, and case law, plus their implications for the future. You could also include writing those codes, statutes, and regulations. This function will not be replaced by AI/ML in our lifetimes.
3. ADVOCACY - Appearing before courts, tribunals, regulators and administrators, arbitrators, and other forums, and even government clerical workers to advocate for clients. This includes the judges themselves and their staff. Obviously this human element will not be replaced.
4. COUNSELING - Listening to and understanding the client’s needs, knowing the pathways, strategies, and tactics to get your client to achieving their goals, and advising them. This ranges from merely being a person to whom the client can vent all the way to the thousands of dollars per hour lawyers who devise strategies to eliminate tax and liability for mega-corporations and billionaires. Almost no human wants to tell their story to a robot. No AI/ML in our lifetime is going to be able to understand what another human’s needs are, ask the right questions and elicit the needed responses, while showing empathy. No AI/ML in our lifetime will have the trust of a CEO or senior manager. There is a fifth function that smart person didn’t consider:
5. “INSURANCE POLICY” - Hiring a lawyer wins you the right to sue them for malpractice if they screw up. Because of this, lawyers are very thorough, cautious, and conservative, often vetting untested ideas and risk-carrying work through many layers of their law firms. Successful legal malpractice suits are rare, but the threat of them and the reputational damage makes legal malpractice itself very rare.
Could a robot replace immigration lawyers? Maybe in the PROCESS aspect of things (filling out the forms). After all, TurboTax is perhaps already as good as an accounting firm for the vast majority of Americans’ taxes, and you could perhaps say the same of Willmaker Pro compared to an estate planning law firm for most Americans. But that seems unlikely, because, unless the mechanical process of filling out a tax form, you have to interview the people, who want a human to listen to them, understand the employer’s needs, and come up with a strategy. The client wants a human to explain it to them and comfort them as to their odds. If the AI/ML screws up, only a human can fix it (low probability) at multiples of the cost of hiring them in the first place. Remember, immigration law ...