BadMoonRisin said:

---

how god dang dumb do we have to get before we see this as a problem?

---

We already see it as a problem.

BMX Bandit said:

---

BadMoonRisin said:

---

how god dang dumb do we have to get before we see this as a problem?

---

We already see it as a problem.

---

Was on a hie this past week with a family that had some high school aged Brit children. I wasn't engaging, just cheating.

They were discussing all the things they try and do to prevent kids from cheating on a device. The students have to hand in their device hen test in little cubbies. If caught with device, it becomes property of the school.

Solution for student…they buy burner phones.

An eventual issue is that the ether will continue to be filled with false, bad, misleading and misconducted information. So an average joe thinking he can save some money on his legal fees is going to use AI to generate lawsuits. Losing barriers to sue other people and particularly businesses (I have a couple of construction businesses) is a bad, bad thing.

I said in earlier discussions there intelligence in AI is artificial. People need to understand it. It is incapable of creativity, intuition and original thought. I don't think it will ever be truly capable of such unless someone has sex and gives birth to an AI computer. Anything else is man made and subject to the terms and conditions put into by said men/women/they/their/bit/byte.

We are already seeing average joes use AI to draft complaints and issue threats. So far they have been laughable, but not for long.

For every incompetent DEI hire exposed, there are 10 out there hiding behind their paychecks.

doubledog said:

---

For every incompetent DEI hire exposed, there are 10 out there hiding behind their paychecks.

---

You are describing "management".

You're onto something here. People are complacent, nihilistic, and less willing to sacrifice for something greater than ever. That's because they've lost their sense of purpose. The thing is, that is God-given. There is indeed virtue in wanting to do things well, in a way that will garner respect rather than ridicule. But the message more and more people are getting is to also do what you enjoy, "live your best life." And more people are coming to the natural conclusion that for them, that involves sticking one's neck out as little as possible.

I rewatched "Prince of Egypt" last night, and I would challenge anyone to claim that that wasn't very well-made, whether someone agrees with the message in it or doesn't. Yes, it was made before AI, or before everything animated was computer-generated. But the reason something like that was made to be so good was because the people involved were inspired to do well. And it is very hard (I would say impossible) to do anything well without being motivated outside one's self.

And here my impression of Georgia lawyers was just that they're really thick-headed and slow to learn. Who knew thst they were also lazy and that they lie a lot?

Quote:

---

And here my impression of Georgia lawyers was just that they're really thick-headed and slow to learn. Who knew thst they were also lazy and that they lie a lot?

---

It's not geographically limited to GA. We are seeing stories all over the US with attorneys getting called out by courts and judges (including federal judges) getting called out for the same thing. Are any actually held to account? Nope.

The various state bars are supposed to be self-policing entities. They really don't do much unless an attorney actually steals from their client - particularly from their escrow accounts. Otherwise? no.

The judiciary is even worse - particularly on the federal side. As we've seen play out over the past 16 months, being a federal judge is never being held to account for anything.

I use AI in my practice. It's helpful for assistance in harmonizing two or more clauses into a single provision. I use it when faced with a question I'd have to ask another lawyer about, but cannot. I use it to get me started on a project if it is a novel concept. I also use it as a final check on everything once a final draft is done.

I do not let it cite legal authority. I may use it as a research assistant but, I always read the case and make sure it stands for what AI says it does.

I feel the laziest about the harmonizing function. I try not to use it too much, but sometimes it's necessary to get something finished. I don't like to use it for this function because I believe you use your ability to think or you lose it.

I just finished my first Google Workflow Studios agent today, and Google just made it available to me this morning. Claude Code is the bomb. I anticipate this agent will save me 5 hours a labor per week.

I'm going to use the extra five hours to create another agent I have mapped out that I think will save me another 5 hours a week. It's a little more complicated, so it might take 10 or so labor hours to get it right, but that will be fine.

Once those are done, I will use the extra time to implement a third project. It will take me about 20 hours of excel work (unless AI can do that for me too, but I doubt it, pretty complex models), and then probably about 5 hours of AI agent work. I think that will save me about 10 hours a week.

I think I will be pretty well set by then. Cut my workload from about 50 per week to about 30. And my hours are cyclical, so I think my schedule will alternate between 40 and 20 hours, week to week. And these are all pretty easy projects.

I'm going to try to find something more meaningful to do with my time than hang out here, but my addition here is a very hungry animal.

Quote:

---

I'm going to try to find something more meaningful to do with my time than hang out here,

---

Impossible!!

She is the appellate lawyer. Would she have been part of the trial court motion opposing a new trial due to the defendant having a lousy lawyer?

Quote:

---

She is the appellate lawyer. Would she have been part of the trial court motion opposing a new trial due to the defendant having a lousy lawyer?

---

She was with the DA's office that prosecuted the Appellant - so she was also trial counsel for the State and she submitted the draft court order to the lower court judge who signed the order with the same false cites as were in the brief for this appeal.

I'm not a litigator but appellate work is a specialized skill. Years so, I met a lawyer with Harris County, Texas and his job was to write briefs for criminal cases being appealed. I don't know if he handled oral arguments. And this is to the GA Supreme Court. I would expect a specialist. That being said, she did state "I did prepare an order. That order was revised.

Its pretty simple really. If you want to use AI for research, create on agent to do the searching and compiling and create another agent to do the review and fact check. AI is pretty good at fact checking as long as you give it the proper instructions on how to fact check.

Quote:

---

I'm not a litigator but appellate work is a specialized skill. Years so, I met a lawyer with Harris County, Texas and his job was to write briefs for criminal cases being appealed. I don't know if he handled oral arguments. And this is to the GA Supreme Court. I would expect a specialist. That being said, she did state "I did prepare an order. That order was revised.

---

She said it was revised - that doesn't mean it's true. How else do you explain phantom cases, presented originally by the DA in prior filings, ending up in the lower court's order? Typically when a judge asks a party to draft an order, they may make a few minor revisions, but the Court is not going to independently verify the case cites that late in the game.

As for your point about appellate advocacy being specialized - it has become that way at certain levels, but plenty trial lawyers still practice in front of their various state Courts of Appeal and higher and when you are talking about government agencies, even more so.

An associate in my office received a brief from opposing counsel that had AI hallucinations. The dude is in big trouble, as he should be. Any lawyer who submits a hallucination to court should be disbarred or severely punished.

I am the 'tech guy" at my firm, and I have been trying out all the different programs. My experience with AI is that six months ago, ChatGPT was bad about hallucinations and correct citations, but the case did not support the proposition advanced. I quit using it. It may be better now.

Then I used Google Gemini and it was better but still had errors.

Westlaw has its version of AI, and it was very good and did not produce any errors that I saw. But I thought it was clunky and their billing terms are insane. They wanted to sign a 4- or 5-year contract. Dude, I don't know if I'm going to have a job in two years, let alone 5. I just signed a 1 year contract with no AI, law only.

Claude Cowork is absolutely phenomenal. My law partner had a very complex inverse condemnation case, and it produced a legal memorandum that was a good as an associate would write at one of the top firms in Atlanta. It even thought of ideas he hadn't thought of. Kinda funny because he's going up against one of my best friends who represents the city. I haven't told him yet that he's fighting AI, not just my 70-year-old partner. Of course we are double checking all citations.

Over the past two weeks, I used Claude Code to create a family touchscreen kiosk in the kitchen. It has a calendar, grocery list, hardware list, and task lists for each family member. It has local weather and we can all access and edit via our phones. I also created a website for my extended family for a house on the Georgia coast we all share. From idea to published and working on the web in about an hour. Amazing.

I have an admin law case with a realtor accused of shenanigans by an ex-business partner. In ten minutes, it pulled out all the relevant facts with dates and documents. It had all the parties listed with phone numbers addresses and emails. It identified documents needed, suggested opening and closing arguments, legal citations and direct and cross examination. I am still checking everything, but this is something I would have to develop over a week of work. it did it in about 15 minutes. Everything I've checked has been spot on.

I have Clio for case management software, and it is getting an AI component. I had an hour zoom conference with Filevine last Friday and another zoom conference at 1:30 pm to discuss billing. Filevine looks amazing. It basically does everything my legal assistant does. Billing summaries, medical summaries, drafting demand letters, etc. I would never fire her, but this will allow her to do more probably about 10X faster.

I think AI is going to be as significant as the industrial revolution. From farming in the field to factory life. The development of the internet is peanuts compared to the social and economic upheaval AI will bring.

Dirty_Mike&the_boys said:

---

I mean seriously, so what can be the answer here?

Paralegals and interns have been doing my work for years but now I used AI and did not check its work?

Really what are the chances that an attorney who is this sloppy, negligent, or dishonest enough to turn in such bad work is going to completely own this without obfuscation or blame-shifting? I'm thinking the chances are pretty good she's not going to own this alone.

---

I have yet to see an incompetent person own up to their mistakes.

Now, an inexperienced person WILL own up to their mistakes. But incompetent? Nope, won't happen.

heres a question for you (and others). what if the case hallucination is correct? does that warrant sanctions? disbarment?

example, lawyer cites a correct proposition of law that is not really disputed, but the case is made up.

" Asking for plausible grounds to infer an agreement does not impose a probability requirement at the pleading stage; it simply calls for enough fact to raise a reasonable expectation that discovery will reveal evidence" Elko v. Kiffin, 380 F.Supp3d 150 (S.D. Tex. 2025).

this is obviously not a good practice and AI cites should always be checked for accuracy, but what punishment does this crime warrant?


on another note, if lawyers aren't using AI in their practice, they are falling way behind their competitors.

BMX Bandit said:

---

heres a question for you (and others). what if the case hallucination is correct? does that warrant sanctions? disbarment?

example, lawyer cites a correct proposition of law that is not really disputed, but the case is made up.

" Asking for plausible grounds to infer an agreement does not impose a probability requirement at the pleading stage; it simply calls for enough fact to raise a reasonable expectation that discovery will reveal evidence" Elko v. Kiffin, 380 F.Supp3d 150 (S.D. Tex. 2025).

this is obviously not a good practice and AI cites should always be checked for accuracy, but what punishment does this crime warrant?


on another note, if lawyers aren't using AI in their practice, they are falling way behind their competitors.

---

The lawyer is telling the judge that (fake) case supports that correct proposition -- big no no. Yes, sanctions, including disbarment are warranted.

In a brief, it is important if you have multiple cases supporting a legal principal rather than one. If you string citations in a brief or use citations like accord, or e.g., see generally. That tells the judge that there are multiple cases supporting this point and it is a well-established legal principle.

The more cases that support a legal principle, the more likely the judge will apply it to your case. It is dishonesty to the Court which, along with stealing your client's money, is about the worst thing a lawyer can do.

https://www.law.georgetown.edu/wp-content/uploads/2022/05/Bluebook-Signals-Explained-FinalJEL-editsFCD.pdf

How much do you charge for the work AI is doing and how? Time spent in ten minute intervals like you would a human clerk or paralegal?

i agree with all you said in principal.



but the quote I gave is from a supreme court case, Twombly its not in dispute that is that is the law. you would never need a string cite for something so basic. you arent trying to convince the judge to apply it to you case, he already is going to be doing so

aggiehawg said:

---

How much do you charge for the work AI is doing and how? Time spent in ten minute intervals like you would a human clerk or paralegal?

---

If you bill anything more than actual time spent, that is unethical and basically theft from the client. You could charge a small fee for what you pay for the AI service. And you should check every citation, which takes time and you could bill for that.

BMX Bandit said:

---

i agree with all you said in principal.



but the quote I gave is from a supreme court case, Twombly its not in dispute that is that is the law. you would never need a string cite for something so basic. you arent trying to convince the judge to apply it to you case, he already is going to be doing so

---

You probably would receive a lesser sanction. But if I'm a judge (in reality, his law clerk) and I look up a citation and it isn't there, I'm going to be pissed off. Is the judge/law clerk going to then do his own research to verify the stated principle is good law? No. They are going to hold you in contempt or refer you to the bar for sanctions. Or dismiss your complaint or answer and then you have to turn the case over to your malpractice insurance.

Best case would be the Judge issues a show cause order, and you have to go to open court, look like a moron in front of everyone there and grovel and beg the judge for forgiveness.

Queso1 said:

---

Doubtful anything serious will happen to her. Some people go through life being the tallest midget. In another thread I addressed the lack of seriousness in our leaders, but this is a problem across society.

I'm not sure if "seriousness" is the right word. Maybe it's that so many people just don't care anymore. I cannot put my finger on it. There's just something missing. It's like the nurses that do the dance videos. Or the politicians ****ing all the time. There's no pride. No philosophy. No attention to style. From the moron store clerk on their phone during customer interactions to this clown - something is off.

It's cultural rot and I am having a difficult time to really define it. It's like I see it, but I can't develop my thoughts on it.

I think one big issue is doing stuff solely for the money or the result. My philosophy is to do stuff for the sake of doing it, not the money…not the perfect product. Those things come from doing it right, but they shouldn't be the point.

I try to mentor my younger employees on this. I tell them, the system is set up where we really don't have to get out of bed in the morning. Why do we do it? Because there is virtue in work. There is virtue in focusing on your work. Phoning it in using AI really irritates something deep down in my life philosophy.

---

It's mass media and communications. People see others getting away with not trying hard or being rewarded for mediocrity or even succeeding and profiting from poor effort or cheating or fraud and over time it wears on even the most conscientious people: why am I trying so hard when lazy bums or underhanded scum can get so much with so little efffort and apparently no shame. We can't use shame and ostracism to enforce standards anymore because freeloaders or slackers can find plenty of peers online to seek solidarity with, and plenty of people with toxic empathy to make excuses for them.

aggiehawg said:

---

How much do you charge for the work AI is doing and how? Time spent in ten minute intervals like you would a human clerk or paralegal?

---

I don't charge for time the AI spends creating it, nor should anyone.

I'll give you a recent example.

had a new case come in where a patient gave notice to sue a nursing home client and requested records. no details on the claim. it is 15,000 pages of records.

in the olden days (six months ago), you'd have a legal nurse consultant review those records, create a timeline, find all the potential liability points, and then let your client know 80 work hours and 6-8 weeks what the case may or may not be about.

I plugged those records into Claude, and got back a 16 page summary of everything I could possibly need to know. Reading that and making my own notes took about a hour which was billed to the client noting that it was AI analysis. then I drafted a letter telling client key liability issues, my own thoughts, and provided those pages of the record, which took about two hours.

Quote:

---

I plugged those records into Claude, and got back a 16 page summary of everything I could possibly need to know. Reading that and making my own notes took about a hour which was billed to the client noting that it was AI analysis. then I drafted a letter telling client key liability issues, my own thoughts, and provided those pages of the record, which took about two hours.

---

You answered my next question about disclosure to the client. I would think that for malpractice reasons as well as ethical ones that disclosure would be mandatory.

Thank you for the response. Glad I am not practicing in this day and age, though.

absolutely needs to be disclosed to them.

BMX Bandit said:

---

aggiehawg said:

---

How much do you charge for the work AI is doing and how? Time spent in ten minute intervals like you would a human clerk or paralegal?

---

I don't charge for time the AI spends creating it, nor should anyone.

I'll give you a recent example.

had a new case come in where a patient gave notice to sue a nursing home client and requested records. no details on the claim. it is 15,000 pages of records.

in the olden days (six months ago), you'd have a legal nurse consultant review those records, create a timeline, find all the potential liability points, and then let your client know in 6-8 weeks what the case may or may not be about.

I plugged those records into Claude, and got back a 16 page summary of everything I could possibly need to know. Reading that and making my own notes took about a hour which was billed to the client noting that it was AI analysis. then I drafted a letter telling client key liability issues, my own thoughts, and provided those pages of the record, which took about two hours.

---

It is amazing. I have no idea what the field of law will be like in two years, or even six months from now.

The problem with Claude is that you have to get Teams so it won't train on your data and theoretically violate attorney client privilege. For HIPAA compliance, you have to get their Enterprise version. Teams has a minimum of 5 members and Enterprise has a minimum of 20 when I last checked.

That's one reason I'm looking to Filevine. They have all the data protection.

I haven't done this demo yet, but they have a real-time AI assistant add on that will analyze the deposition testimony for your real time at the deposition and suggest further cross examination questions. That's insane.

Quote:

---

I haven't done this demo yet, but they have a real-time AI assistant add on that will analyze the deposition testimony for your real time at the deposition and suggest further cross examination questions. That's insane.

---

We have been evaluating in-house tools, but they are changing so fast, evaluations are difficult.

I'm concerned we are taking away human judgement at a speed we cannot absorb.

And we are constructing unverified junk or questionable records of AI output that future AI may naively reference at an astonishing rate.

We'll need to clearly label or tag AI work product and error check it with other AI regularly

Aggie Jurist said:

---

Quote:

---

I haven't done this demo yet, but they have a real-time AI assistant add on that will analyze the deposition testimony for your real time at the deposition and suggest further cross examination questions. That's insane.

---

We have been evaluating in-house tools, but they are changing so fast, evaluations are difficult.

I'm concerned we are taking away human judgement at a speed we cannot absorb.

---

It is very weird to me but particularly as it relates to this case. This case involved an older statute that has since been repealed in full. (Citizen's arrest no longer allowed after Arnaud Arbury case.) And would AI account for the subsequent repeal? Could that be part of the confusion?

Consider that the dumbest people you know are repeatedly being told "You're absolutely right!" by LLMs.

— Jameson Lopp (@lopp) March 23, 2026

even Westlaw's AI gets case information wrong

More lawyers misusing AI (Westlaw's CoCounsel):

6th Cir. opinion today focuses on criminal defense attorney who "used artificial intelligence to draft the briefs in this case and then filed them without properly verifying the cited legal authorities."

"The first tell" that AI… pic.twitter.com/6Z70CIq5sZ

— Rob Freund (@RobertFreundLaw) April 3, 2026

CanyonAg77 said:

---

So, how likely is it that these fake cases came from Google or AI, and no one checked them?

---

100%

Someone didn't realize that AI will absolutely make up an answer if it can't find a real one, just to be "helpful".