How often do you use AI in your job?

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:As a lawyer what is the best AI tool for summarizing video meetings without creating privileged issues? That I think would actually be useful but not sure safe to adopt given the data issues.

Any LLM can summarize a meeting transcript.

I’m looking for one that will also create the transcript.

Check out Fireflies.

But I think every AI tool will arguably create a potential privilege issue- is disclosing it to the Ai tool disclosure to a 3rd party?? - although that’s just more work for the lawyers!

Again, this is a piece of software. Your company should have a closed version.

True but I’m just saying some crafty lawyer or class action lawyer will eventually try to find an argument around it.

But yes, a lot of the concerns voiced here are addressed by paying the $ for an enterprise version for your co

Um. No. Closed software is closed software. Your “crafty lawyer” has no case unless the companies screw up.

Sigh, you must be a dude, right? I’m not sure what lingo you’re trying to use but most SaaS tools are a combo of closed and open software. I assume you’re trying to say the info is secure and isn’t shared, but honestly I think you’re just throwing around lingo you don’t fully understand. You should be calling out SOC and encryption requirements but whatever…

The point is that you’d be surprised at what behind the scenes tracking that lawyers are eventually able to find to bring claims under awkwardly written laws that don’t translate into modern uses

[Anonymous]

Screenwriter. Use it to generate ideas, clean up text, detect redundancy, generate plot lines. Explore alternatives. Streamlines some "left" brain work so I can focus on the more emotive, "right" brain work. It's quite fun.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

[Anonymous]

I used it to generate images for a presentation recently and it was great! Will never be using clip art again.

I am a manager and use it often for performance reviews, creating task lists, writing emails, turning emails in to bullet points, proofreading documents, preparing a business case memo, formatting agendas, making tables, creating schedules.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:As a lawyer what is the best AI tool for summarizing video meetings without creating privileged issues? That I think would actually be useful but not sure safe to adopt given the data issues.

Any LLM can summarize a meeting transcript.

I’m looking for one that will also create the transcript.

Check out Fireflies.

But I think every AI tool will arguably create a potential privilege issue- is disclosing it to the Ai tool disclosure to a 3rd party?? - although that’s just more work for the lawyers!

Again, this is a piece of software. Your company should have a closed version.

True but I’m just saying some crafty lawyer or class action lawyer will eventually try to find an argument around it.

But yes, a lot of the concerns voiced here are addressed by paying the $ for an enterprise version for your co

Um. No. Closed software is closed software. Your “crafty lawyer” has no case unless the companies screw up.

Sigh, you must be a dude, right? I’m not sure what lingo you’re trying to use but most SaaS tools are a combo of closed and open software. I assume you’re trying to say the info is secure and isn’t shared, but honestly I think you’re just throwing around lingo you don’t fully understand. You should be calling out SOC and encryption requirements but whatever…

The point is that you’d be surprised at what behind the scenes tracking that lawyers are eventually able to find to bring claims under awkwardly written laws that don’t translate into modern uses

No. I’m not throwing around “lingo” I don’t understand. I work for an org that simply has a closed instance of GPT-40. It’s not that complicated to have a closed instance of LLM software. You don’t need to understand every element of cybersecurity to know what an agreement not to send the data back means.

No one is suing over Microsoft Teams or Word, which also have access to all your private corporate data. In the end that is how this ends up. No one is going to sue over it unless these companies are really screwing around which in my opinion is not worth it when people are providing so much testing data on free instances of Chat-GPT.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

DP and I think your own assumptions are keeping you from appreciating the utility of the software.

First of all the media is the one insisting that AI is going to outright replace humans when we have all seen that LLMs are simply not capable of doing that. They are just a tool.

That doesn’t mean they are “garbage.” That is like saying that Google search is garbage because it can’t write a legal brief.

I will give you a simple example: our corporate version of westlaw gave us Cocounsel. I had an annoying task where I had to look up a certain law for every state. It was able to give me an easy link for that in 1 go.

No, that’s not writing for me but it is saving me time and it is “smarter” than the regular search function. That’s enough to increase my personal productivity. I’ve been able to find very specific clauses in documents using the Adobe AI search tool much faster than I ordinarily would because I can ask it for the meaning, not just the exact phrase.

I simply don’t understand how people are saying these tools are not helpful, not saving them time, etc. If you experiment and are creative, the personal and professional uses seem so unlimited. It’s the most genuine fun I have had at work in a long time.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

DP and I think your own assumptions are keeping you from appreciating the utility of the software.

First of all the media is the one insisting that AI is going to outright replace humans when we have all seen that LLMs are simply not capable of doing that. They are just a tool.

That doesn’t mean they are “garbage.” That is like saying that Google search is garbage because it can’t write a legal brief.

I will give you a simple example: our corporate version of westlaw gave us Cocounsel. I had an annoying task where I had to look up a certain law for every state. It was able to give me an easy link for that in 1 go.

No, that’s not writing for me but it is saving me time and it is “smarter” than the regular search function. That’s enough to increase my personal productivity. I’ve been able to find very specific clauses in documents using the Adobe AI search tool much faster than I ordinarily would because I can ask it for the meaning, not just the exact phrase.

I simply don’t understand how people are saying these tools are not helpful, not saving them time, etc. If you experiment and are creative, the personal and professional uses seem so unlimited. It’s the most genuine fun I have had at work in a long time.

This is true. I think I’m realizing though how infrequently my job has drudge tasks to do like the ones you have listed. If I had more of them, I’d use AI.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

DP and I think your own assumptions are keeping you from appreciating the utility of the software.

First of all the media is the one insisting that AI is going to outright replace humans when we have all seen that LLMs are simply not capable of doing that. They are just a tool.

That doesn’t mean they are “garbage.” That is like saying that Google search is garbage because it can’t write a legal brief.

I will give you a simple example: our corporate version of westlaw gave us Cocounsel. I had an annoying task where I had to look up a certain law for every state. It was able to give me an easy link for that in 1 go.

No, that’s not writing for me but it is saving me time and it is “smarter” than the regular search function. That’s enough to increase my personal productivity. I’ve been able to find very specific clauses in documents using the Adobe AI search tool much faster than I ordinarily would because I can ask it for the meaning, not just the exact phrase.

I simply don’t understand how people are saying these tools are not helpful, not saving them time, etc. If you experiment and are creative, the personal and professional uses seem so unlimited. It’s the most genuine fun I have had at work in a long time.

This is true. I think I’m realizing though how infrequently my job has drudge tasks to do like the ones you have listed. If I had more of them, I’d use AI.

What drudge tasks? You never have to look up laws? You don’t have to look up information in documents? You don’t have to study proposed legislation? You absolutely never have to create presentations or trainings or blog posts? You never have to manipulate data or extrapolate the result of policy choices? This isn’t drudge, it’s just… part of the work, and it’s not even that unpleasant.

Your entire job is “deep” thinking? You never have to save time? What does that even entail?

[Anonymous]

Anonymous wrote:

Anonymous wrote:I work with PII as a government attorney. Absolutely not permitted to feed that into AI. So never.

So you are naturally a scared person …

Lol. No. I’d be violating the law if I fed PII into AI.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I work with PII as a government attorney. Absolutely not permitted to feed that into AI. So never.

So you are naturally a scared person …

Lol. No. I’d be violating the law if I fed PII into AI.

Why would you feed PII into AI? You can use it without doing that. Most government agencies have their own approved LLMs at this point. OpenAI offered 1 year of gpt-40 to agencies for $1. You’re supposed to develop some familiarity with it.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

DP and I think your own assumptions are keeping you from appreciating the utility of the software.

First of all the media is the one insisting that AI is going to outright replace humans when we have all seen that LLMs are simply not capable of doing that. They are just a tool.

That doesn’t mean they are “garbage.” That is like saying that Google search is garbage because it can’t write a legal brief.

I will give you a simple example: our corporate version of westlaw gave us Cocounsel. I had an annoying task where I had to look up a certain law for every state. It was able to give me an easy link for that in 1 go.

No, that’s not writing for me but it is saving me time and it is “smarter” than the regular search function. That’s enough to increase my personal productivity. I’ve been able to find very specific clauses in documents using the Adobe AI search tool much faster than I ordinarily would because I can ask it for the meaning, not just the exact phrase.

I simply don’t understand how people are saying these tools are not helpful, not saving them time, etc. If you experiment and are creative, the personal and professional uses seem so unlimited. It’s the most genuine fun I have had at work in a long time.

This is true. I think I’m realizing though how infrequently my job has drudge tasks to do like the ones you have listed. If I had more of them, I’d use AI.

What drudge tasks? You never have to look up laws? You don’t have to look up information in documents? You don’t have to study proposed legislation? You absolutely never have to create presentations or trainings or blog posts? You never have to manipulate data or extrapolate the result of policy choices? This isn’t drudge, it’s just… part of the work, and it’s not even that unpleasant.

Your entire job is “deep” thinking? You never have to save time? What does that even entail?

i mean, if i have to look up information in documents... i look at the document? i don't ask someone else what they think was important in the document. my entire job is to know things and be accurate. checking the output of AI is a lot more tedious than just understanding something and writing it myself.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

DP and I think your own assumptions are keeping you from appreciating the utility of the software.

First of all the media is the one insisting that AI is going to outright replace humans when we have all seen that LLMs are simply not capable of doing that. They are just a tool.

That doesn’t mean they are “garbage.” That is like saying that Google search is garbage because it can’t write a legal brief.

I will give you a simple example: our corporate version of westlaw gave us Cocounsel. I had an annoying task where I had to look up a certain law for every state. It was able to give me an easy link for that in 1 go.

No, that’s not writing for me but it is saving me time and it is “smarter” than the regular search function. That’s enough to increase my personal productivity. I’ve been able to find very specific clauses in documents using the Adobe AI search tool much faster than I ordinarily would because I can ask it for the meaning, not just the exact phrase.

I simply don’t understand how people are saying these tools are not helpful, not saving them time, etc. If you experiment and are creative, the personal and professional uses seem so unlimited. It’s the most genuine fun I have had at work in a long time.

^ This is the kind of AI promotion that makes me frustrated and, frankly, suspicious. Why are you so invested in other people using this tool?

Other PPs have said they tried it and it's not helpful for their jobs. It's weird that the AI fans can't just accept that at face value. If PPs are wrong ... why do you care?

IME, the people who react like this to other people saying "no thanks" are either selling something, or really insecure in their choices and need the validation of "everybody's doing it" to paper over the problems with what they choose.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

You’re jumping ahead, and clearly not as plugged into it as you think you are. No one is saying you put in a prompt and get an output and run with it exactly as is. But it gets you very far, in mere seconds. You’re already using AI of course and dont even really think about it. Every apple and google product uses AI extensively in both front end and back end processes, and has been for years. But we are talking about GAI here. Generative.

As someone said above, ignore this at your own peril.

As far as law firms, I’m the in house lawyer poster above who said that I’m calling outside counsel less now bc I can do certain things on my own now.

As I one example of it’s capabilities, a 2k an hour outside firm (with multiple lawyers on the task) was asked to draft something for my co. It took them a week and a half to get us a draft, and it wasn’t great. Not entirely their fault, our directions weren’t great and they came back with something far too detailed for us to use. Of course setting up time with their team and my co to talk again takes time, and other things got in the way, so we are still waiting on it.

As an experiment, that very first day, I asked chat for the same document using the same parameters and prompts. I had to tailor my request a few times- and admittedly that takes a human hand with a knowledge base- but within 15 minutes I had a document that was better than what the outside counsel produced.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

DP and I think your own assumptions are keeping you from appreciating the utility of the software.

First of all the media is the one insisting that AI is going to outright replace humans when we have all seen that LLMs are simply not capable of doing that. They are just a tool.

That doesn’t mean they are “garbage.” That is like saying that Google search is garbage because it can’t write a legal brief.

I will give you a simple example: our corporate version of westlaw gave us Cocounsel. I had an annoying task where I had to look up a certain law for every state. It was able to give me an easy link for that in 1 go.

No, that’s not writing for me but it is saving me time and it is “smarter” than the regular search function. That’s enough to increase my personal productivity. I’ve been able to find very specific clauses in documents using the Adobe AI search tool much faster than I ordinarily would because I can ask it for the meaning, not just the exact phrase.

I simply don’t understand how people are saying these tools are not helpful, not saving them time, etc. If you experiment and are creative, the personal and professional uses seem so unlimited. It’s the most genuine fun I have had at work in a long time.

^ This is the kind of AI promotion that makes me frustrated and, frankly, suspicious. Why are you so invested in other people using this tool?

Other PPs have said they tried it and it's not helpful for their jobs. It's weird that the AI fans can't just accept that at face value. If PPs are wrong ... why do you care?

IME, the people who react like this to other people saying "no thanks" are either selling something, or really insecure in their choices and need the validation of "everybody's doing it" to paper over the problems with what they choose.

I’m a lawyer and it seems like many of the anti AI people on here are lawyers (btw it’s GAI we are discussing, the AI ship sailed long ago and we are all ‘using’ it indirectly) and it’s a little embarrassing to be part of a group that looks so boomer and out of touch. Many other lawyers in my co were similarly uncertain about GAI but we had to get on board with its use by at least the rest of our company - and find ways to put guardrails up. It actually wasn’t that difficult.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Lawyer- I use it to help me re write emails. It helps with my tone. I’m female and have always struggled with being too nice and taking on too much work. And then I swung to being too rude in emails. AI gives a good balanced middle tone.

The lawyers in this thread who are being so negative about AI simply don’t understand how vast it can be in its uses. It is such an awesome tool even if it can’t do a “legal” analysis. It can do the first draft of something like a blog post, it can create meeting minutes from a transcript, it can create detailed notes from a transcript, it can create a full PowerPoint presentation from a compliance document, it can help turn text into tables and develop databases, it can help create really advanced excel tools to do analyses, it can teach you how to create advanced templates. It’s amazing.

Np also a lawyer who has barely used ai. Our admin people are being encouraged to use it, I can see it would be helpful for them. Lawyers by and large are not for the reasons mentioned on this thread.

Lawyers on dcum typically are the best of the best lawyers - we graduated top of our top law schools and trained in some of the most demanding work places in the US. People like to shit on lawyers and say book smart doesn’t mean smart, but reality is that to be an attorney in my firm you need to be exceptionally bright, hard working, productive and good at stuff. You just didn’t get through college and law school with top grades without being so.

With that background, all the things you list…. I do perfectly and efficiently the first time. Other people on this thread saying it takes them “a quick” twenty mins to draft a very short email? My short emails take 2-3 mins and while I concede may have a minor typo, their tone and content is flawless. That’s why my clients pay me $2000 an hour. I need to be able to send that kind of client product for 10 hours a day in near consistent quality. I can produce the email, structure the excel, draft the article, etc in perfect form the first time in the same time it would take me to input the information in ai. Certainly less time than it would take me to review and edit the ai output. And the ai output is likely to be just not as good as what I can do. If I could not do it that well and that quickly the first time, I would have been pushed out of biglaw a long time ago.

I have friends in regular non legal jobs and their jobs are just less demanding. The hours and deadlines are less demanding and the work output is less demanding. My dh is a non lawyer (and very bright and very successful) but he may spend all day just debating a draft email he needs to send to the ceo, and he can get away with that in his job. I could see how ai might add value in that context (although even then, the reason why dh makes seven figures for a job where his only task in a day is writing an email is because he is damn good at communication and very well liked, so the times we’ve put his emails thru ai we’ve been very disappointed with the output because it no longer sounded like him).

You say your emails take 2-3 minutes… why would I pay you $2000 if an AI and a “less smart” person will soon also take 2-3 minutes to do the same thing? You’re the one in the most danger.

Not the PP you’re quoting but this is a very dumb takeaway. Those 2-3 minute emails are not the core of our job. It’s the writing and research and analysis that AI can’t do. The email is a short summary or update of the real work that AI can’t do.

Of course it can’t do your job (yet). But what if it can do 20% of your job? 40? 60? When do you get devalued?

Here is the reality. The top flight lawyers in this thread are creating exactly the content AI is training on to TRY to SOUND like us. Not to actually reason or analyze like us. To SOUND like it is. It’s not even coming close to the actual work. It’s just rhetorical cosplay. Looks like the real thing at first glance, but is useless garbage on closer review.

You are anthropomorphizing the AI the same way its boosters would. It’s not a person. It’s not a dumb assistant who is never going to be as smart as you. It is a program that converts words to numbers and gets a statistical approximation for the next word. It doesn’t ever need to be as good as you. It needs to be close, and for a person who is cheaper than you to get the rest of the way there.

No, I’m not. I’m describing how it works. It trains on the kind of work we do, to sound like we sound. Not to do the work we do, but to sound like it did. If you think my work is to *sound* like a lawyer, you misunderstand the task.

Years ago, people insisted that software would never be able to provide accurate translation in real time. Today I was in a cab in Athens communicating with the Greek driver via translate apps on our phone.

Ignore what is happening at your own peril.

Look. I wish it did what people say it does. I’m GC at a tiny nonprofit. If it did even 10% of what people claim, I’d be thrilled! And if it ever does get to the point it can replace a team of lawyers, I’ll be ruling the world! But as it stands currently it’s literally useless to me. I don’t need help “summarizing a long email” as someone upthread said. I don’t need window dressing or help changing the tone of my writing. I need major substantive legal analysis and strategy.

Your world is very small, and you need to open it. Wake up.

It is not 100 percent there on analysis and strategy, but it’s about 50+ % there in many areas.

But it just isn’t. If this were true, orgs like mine and elite boutique law firms would already be stopping hiring associates and find themselves able to take on more cases. That’s not even close to what’s happening. I’m very plugged into all this. There are many of us who, as I say, are ready and willing to take on more work if these tools really were 50% of the way there. But they are really and truly garbage. They sound good on a surface level but produce trash you can’t use once you actually look closely.

DP and I think your own assumptions are keeping you from appreciating the utility of the software.

First of all the media is the one insisting that AI is going to outright replace humans when we have all seen that LLMs are simply not capable of doing that. They are just a tool.

That doesn’t mean they are “garbage.” That is like saying that Google search is garbage because it can’t write a legal brief.

I will give you a simple example: our corporate version of westlaw gave us Cocounsel. I had an annoying task where I had to look up a certain law for every state. It was able to give me an easy link for that in 1 go.

No, that’s not writing for me but it is saving me time and it is “smarter” than the regular search function. That’s enough to increase my personal productivity. I’ve been able to find very specific clauses in documents using the Adobe AI search tool much faster than I ordinarily would because I can ask it for the meaning, not just the exact phrase.

I simply don’t understand how people are saying these tools are not helpful, not saving them time, etc. If you experiment and are creative, the personal and professional uses seem so unlimited. It’s the most genuine fun I have had at work in a long time.

^ This is the kind of AI promotion that makes me frustrated and, frankly, suspicious. Why are you so invested in other people using this tool?

Other PPs have said they tried it and it's not helpful for their jobs. It's weird that the AI fans can't just accept that at face value. If PPs are wrong ... why do you care?

IME, the people who react like this to other people saying "no thanks" are either selling something, or really insecure in their choices and need the validation of "everybody's doing it" to paper over the problems with what they choose.

Um… ok? You can be as “suspicious” as you like.

I think you are reading way too much into an online discussion. Someone is saying “this tool is useless.” Someone else is saying “this tool is extremely useful for x, y, z,” and the response to that I am seeing is “I am too smart/highly paid for this to be useful for me. It must be useful to you because all you do is drudgery.” That’s wrong in addition to being rude, and it’s a very weak argument against a technology that is as transformative as AI has already proven itself to be.

The LLMs you dabble with at work are not all the AI around you- it’s already had profound impacts on medical applications, GPS, translation software, etc. This is going to impact your career, whether you are in big law or government or the nonprofit sector. I don’t actually care whether you believe that. Time will tell us who is right.