The most prestigious law school admissions discussion board in the world.
|
 |
 
|
Braggarts claiming AI makes them more productive genuinely scare me
Poast new message in this thread
Date: April 19th, 2026 11:08 AM Author: cracking goyim plaza these are the snake oil salesmen of our era (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827078) |
Date: April 19th, 2026 11:15 AM Author: adventurous stock car haunted graveyard It makes me meore productive. Not a game changer but definitely helpful. Like a first year associate that can do basic research that you have to double check, but it's still good to have. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827094) |
 |
Date: April 19th, 2026 11:19 AM Author: Impressive parlour cr.
i dont understand labeling it as "bragging." it's like saying people who talked about computers making office work more efficient are "bragging." it's just a tool that some people find very useful. how is that "bragging?" |
|
Date: April 19th, 2026 11:42 AM Author: pungent internet-worthy locus Mid-wits get MAF about anything AI related. It's that simple. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827129) |
|
Date: April 19th, 2026 12:00 PM Author: Impressive parlour it's the midwit millennial version of the boomer check thing poasted below.
interestingly, boomers think ai is magic. |
|
Date: April 19th, 2026 12:04 PM Author: pungent internet-worthy locus I believe that there are boomers who think it's magic, I can picture that being a basin of attraction in the boomersphere. But the boomers I know all reject it almost outright. The millennial midwits construct convoluted MAF screeds about it filled with all kinds of cope tells. The boomers I know either avoid AI altogether and refuse to use it. Or they use it once in a while then complain and yell about it when it gives them the "wrong answer", then brag about how they "just had to solve it myself because the AI isn't smart enough". (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827155) |
|
Date: April 19th, 2026 12:08 PM Author: Impressive parlour true. but i do find boomers tend to be more receptive, paradoxically. of course there are still plenty of luddite boomers out there too.
i think the reason is millennials are much more aware of technological progress than boomers. the span of time from internet to ai is essentially our entire lives, whereas for boomers it's only like the last third. they're still figuring out how to open pdf, so something lot chatgpt is sometimes indistinguishable from sorcery to them. but yeah, you also have the luddite boomers who bitch about every new tech thing. |
|
Date: April 19th, 2026 12:13 PM Author: cracking goyim plaza It's not about being a luddite, it's all about how the "AI narrative" is just a bunch of bullshit. It's promises without delivery. AI can help you proofread a pleading, but it doesn't know what it's doing. It makes mistakes all the time. It requires you to go back through and redo work. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827176) |
|
Date: April 19th, 2026 12:49 PM Author: pungent internet-worthy locus I see people say this sort of thing a lot but they never give specific examples (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827216) |
|
Date: April 19th, 2026 9:48 PM Author: Supple Masturbator Reading Party Why do you need examples? Everyone knows it still hallucinates. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828278) |
|
Date: April 19th, 2026 10:05 PM Author: pungent internet-worthy locus lol I think its funny how when an ape makes a mistake (which is almost always) noone flinches, but when AI does "hallucination" language gets invoked (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828330) |
|
Date: April 19th, 2026 11:24 AM Author: cracking goyim plaza If that's how you're using it, I wouldn't call that bragging. What I mean by bragging is someone using AI to make a new LinkedIn post every day about using Claude to apps in python to crawl through case files and how they're implementing AI at every level of their organization and yadda yadda yadda. For me, that's not an effective pitch. Sounds to me like you're more interested in being a data scientist than you are being a lawyer. I want a lawyer. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827113) |
|
Date: April 19th, 2026 11:25 AM Author: Gold histrionic prole lol wtf are you talking about
weird boomer schtick reminiscent of the "put a check in the mail like a grownup sweety" post |
|
Date: April 19th, 2026 11:43 AM Author: cerise talking dilemma I think a lot of these poasts are stealth marketing (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827130) |
|
Date: April 19th, 2026 1:44 PM Author: pungent internet-worthy locus Those are all just lawyers with tech envy that feel powerful now that they can use Scripting (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827265) |
|
Date: April 19th, 2026 9:53 PM Author: ruddy gaming laptop persian It makes me more efficient, but instead of working the same amount of time and producing more, I just nap and fuck around more (not a lawyer) and my output stays the same (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828296) |
|
Date: April 19th, 2026 12:28 PM Author: big wine faggot firefighter garrison (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827191) |
|
Date: April 19th, 2026 12:48 PM Author: pungent internet-worthy locus (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827214) |
|
Date: April 19th, 2026 1:22 PM Author: Impressive parlour (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827239) |
|
Date: April 19th, 2026 1:26 PM Author: Pea-brained vivacious ratface It's amazing for doing initial legal research on a subject. Often you find it's all correct and barely needs to be supplemented at all. Sometimes it has mistakes to some degree, so you're risking a malpractice lawsuit by relying on it without your own ability to doublecheck it.
Have yet to encounter outright hallucinations with the newest models, though I know others have. |
|
Date: April 19th, 2026 1:42 PM Author: pungent internet-worthy locus This is the cr objective description (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49827262) |
|
Date: April 19th, 2026 9:50 PM Author: Supple Masturbator Reading Party Ask it about a line in a movie you’re watching. Ask what the character meant when they said it. Every time I do this, it hallucinates something. It won’t just say idk (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828287) |
|
Date: April 19th, 2026 10:07 PM Author: pungent internet-worthy locus Thats what's funny is that something like "a line from a movie I'm watching" isn't something I would ever ask it and I can't believe there are people who care about this sort of thing. Asking it "what a character meant" and I guarantee you it gives a better answer than an ape would (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828335) |
|
Date: April 20th, 2026 1:22 AM Author: aqua concupiscible blood rage organic girlfriend Some of the sanctions that the federal bench has been doling out for AI mistakes have been draconian, far beyond what the same issue would have gotten you 5 years ago.
The thing's inability to say idk or give a confidence level or caveats is grating on its best day and a serious problem on its worst. |
|
Date: April 20th, 2026 1:05 AM Author: obsidian lay Law is one of the things AI is worst at and it's still somewhat useful. But it's great for code. You should absolutely be using it to write code for you. If you're a typical lawfag, then I guarantee there's some tedious process you still do that should be a python script. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828589) |
|
Date: April 20th, 2026 1:09 AM Author: pungent internet-worthy locus It's because law is more Intellectually Demanding than solving erdos problems (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828591) |
|
Date: April 20th, 2026 1:19 AM Author: obsidian lay It's because law has more faggy sophistry and internet retards writing about it. Nobody puts out fake python code for fun, but every shitlib will write a ten paragraph essay along the lines of "The law says (latest shitlib talking points)." The slop is in the training data. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828599) |
|
Date: April 20th, 2026 2:03 AM Author: pungent internet-worthy locus Yeah that's a big part of it (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828642) |
|
Date: April 20th, 2026 10:00 AM Author: Impressive parlour (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49829004) |
|
Date: April 20th, 2026 1:24 AM Author: grizzly very tactful pistol Explain something that you would use a python script for and how to implement this. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828606) |
|
Date: April 20th, 2026 1:33 AM Author: obsidian lay Python is now easy to write and turing complete. Anything that can be done with a computer can be done with python. It would be quicker to write out a list of things that CAN'T be done in Python (which is isomorphic with anything that can't be done with a computer, like determine the true meaning of love or find the last digit of pi). There are some practical limitations because the tradeoff for python's power and simplicity is that it's a high level interpreted language, so low level compiled languages will generally be faster, but computer chips are so good these days that it doesn't matter, for the tasks lawfags do IRL they won't even notice the performance hit.
There must, for instance, be SOME sort of data you receive and need to understand as part of your job. You should be doing so with specialized code written in panda or polars (depending on the situation). Most lawfags are still derping around in Excel, which is the less powerful tool and they don't know how to use properly anyway. |
|
Date: April 20th, 2026 2:12 AM Author: pungent internet-worthy locus Python actually has libraries just for Law. Eyecite and Juriscraper I think are two of them. Juriscraper lets you scrape opinions and oral arguments or whatever from systems like pacer. You can also use NLP libraries for doc review tasks. And then pandas for shit you'd use excel for it can handle like tens of millions of rows. Probably some outside the box uses too like using sci-kit learn ml to process judge analytics and predict if you'll win your cases. For just small tactical scripts idk what you'd use them for depends if you have a lot of annoying manual tasks. (http://www.autoadmit.com/thread.php?thread_id=5858366&forum_id=2Reputation#49828646) |