How will artificial intelligence affect the legal industry?

[aurora borealis]

On 1/24/2023 at 2:15 PM, realpseudonym said:

It's basically just a plagiarism machine. It doesn't cite sources and doesn't credit anyone else for their work. 

There's some evidence that it's also just making things up:

 

[BHC1]

To follow up on aurora borealis' point above, ChatGPT has no problem confidently reciting completely incorrect information:

Q: Can you provide examples of roadside stop cases where voluntariness of a statement was at issue?

A: Certainly, there have been several cases in Canada where the voluntariness of statements made during a roadside stop has been at issue.

One example is the case of R. v. Grant, where the Supreme Court of Canada considered the voluntariness of a statement made by the accused during a roadside stop. In that case, the accused was stopped by police while driving his vehicle. The police officers were investigating a shooting that had occurred in the area, and they asked the accused a number of questions about the incident. The accused made a statement to the police, in which he admitted to being present at the scene of the shooting. The statement was used as evidence against him at trial, but the accused argued that it should be excluded because it was involuntary. The Supreme Court of Canada held that the statement was involuntary and should be excluded from evidence. The Court found that the statement was the result of the accused's being detained in a coercive environment, where he was surrounded by several armed police officers, and where the police had not advised him of his right to remain silent or his right to retain counsel.

Another example is the case of R. v. Suberu, where the voluntariness of a statement made during a roadside stop was also at issue. In that case, the accused was stopped by police while driving his vehicle. The police officers suspected that he was driving under the influence of alcohol, and they asked him to perform a number of field sobriety tests. The accused failed the tests, and the police arrested him for impaired driving. During the arrest, the accused made a statement in which he admitted to drinking several beers earlier in the evening. The statement was used as evidence against him at trial, but the accused argued that it should be excluded because it was involuntary. The Supreme Court of Canada held that the statement was voluntary and should be admitted as evidence. The Court found that the accused was not in custody at the time he made the statement, and that he had not been subjected to any form of duress or improper influence.

It's important to note that these cases are examples and each one depends on its own set of facts. And that the voluntariness of a statement is determined on a case by case basis and will depend on the specific circumstances of each case.

Any first year law student would be able to tell you that the fact summaries provided above are completely inaccurate, after reading just the case headnotes. AI chat bots will probably become useful tools for lawyers, but they are currently comically bad. 

 

 

 

[Rusty Iron Ring]

A friend of mine asked it a legal question recently, and it provided a number of authoritative cases that, as far as I can tell, do not exist.  They were provided complete with citations. Looking up those citations draws totally different cases on different issues. 

[CleanHands]

ChatGPT is so unbelievably terrible at absolutely everything except formulating plausible-sounding (but completely wrong) sentences that it is shocking, hilarious, and has completely changed my view of artificial intelligence and made me convinced it will be far less of a disrupter for the foreseeable future than I had envisioned.

Saw this on reddit:

On a horrifying note though: https://www.vice.com/en/article/k7bdmv/judge-used-chatgpt-to-make-court-decision 

Edited February 5, 2023 by CleanHands

[CleanHands]

 

[GreyDude]

My browser logged me out of the CanLawForum today, and as I was signing back in I saw the ad in the screenshot below. Has anybody tried it out? I notice from the website that it uses the GPT engine.

On the accuracy of ChatGPT (raised by @aurora borealis, @CleanHands and others), at the start of the semester I asked it to summarize the main theory presented in chapter 5 of H.L.A. Hart's The Concept of Law, since I might have students writing essays on it for a philosophy of law class I teach. The chatbot got the chapter quite wrong by treating one element of the theory as though it were the whole thing. I wish I had a screen capture. If one of my students were to give the same answer on an exam, they would receive part marks but not a passing grade.

To me, the upshot is that the chatbot can deftly and impressively reorganize text in a way that sounds plausible and might be accurate, but it cannot show actual understanding, and thus cannot be trusted if actual understanding is what is needed. I would not trust it if I really needed to learn something and be sure that what I was learning was correct (never mind it's tendency to just make stuff up sometimes).

So I would worry about its use in situations where its output really matters.

Edited March 11, 2023 by GreyDude
clarity