Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
This Is the First Episode There Were No Other Episodes
218 Plays1 year ago
The titular minds discuss AI, its use in the law, ChatGPT, and the story of an attorney being banned from Radio City Music Hall because the owner is a big dope.
- ICYMILaw (https://icymilaw.org/) - Awesome bot running on esq.social aggregating links and generating AI summaries in *podcast* form. Thanks, David (@[email protected])!
- ChatGPT (https://chat.openai.com/)
- Madison Square Garden Uses Facial Recognition to Ban Its Owner’s Enemies (https://www.nytimes.com/2022/12/22/nyregion/madison-square-garden-facial-recognition.html)
- Vanna White Decision (https://h2o.law.harvard.edu/cases/4825)
- Steam Deck (https://store.steampowered.com/steamdeck)
- Don’t Embarrass Trump Over Tax Records Release—Charge Him Instead (https://news.bloombergtax.com/tax-insights-and-commentary/dont-embarrass-trump-over-tax-records-release-charge-him-instead)
Recommended
Transcript
00:00:02
Speaker
Okay.
Introduction to Esquiring Minds Podcast
00:00:03
Speaker
It didn't boot me. Welcome to episode one of Esquiring Minds. It's a podcast where it's just three lawyer friends goofing around for your enjoyment, touching on some legal topics. And none of this for many reasons should be taken as legal advice. I'm one of the three.
Meet the Hosts: Andrew, Jake, and Jason
00:00:19
Speaker
I'm Andrew Leahy. I'm a tax and technology attorney in New Jersey. And I have here Jake. Jake Schumer. I am a Florida local government and land use attorney, former public defender.
00:00:33
Speaker
And I'm Jason Ramsland. I am an employment lawyer. I sue people's bad bosses day and night night. You're working. I'm doing this at night. Yeah. Yeah. Occasionally. Sometimes I said emails to send at night.
Email Etiquette: Nighttime Communication
00:00:49
Speaker
So it looks like I'm burning the midnight oil. I did that last night or because I didn't want to hear back from anybody in response to that email that day.
00:00:57
Speaker
I always feel weird about sending an email at like 10 p.m. and I'm like, oh, I know some people are gonna check their phones and they're gonna get stressed out about this. But I'm working, I'm working sadly, unhealthily, I'm working.
00:01:15
Speaker
And the whole idea behind email was supposed to be, it's supposed to be like asynchronous, right? So you're supposed to be able to just send it whenever, but that convention has completely fallen by the wayside. And I have the same feeling. I'm afraid to send emails to people, you know, past a certain hour or too early in the morning. Yeah. And I'm too lazy to schedule it on Outlook to be like, send this at a normal time, send this at 8am. So it looks like I woke up early.
00:01:37
Speaker
I have no sympathy for people who are going to read their emails at bedtime and get stressed out because the tools are there. People just use the tools that are available to you. You know how easy it is to set up a focus mode on your phone so that your email doesn't come in? You can even hide the app so that you don't even see your email. You can't see your email on your home screen on your phone. No sympathy.
00:01:57
Speaker
So I only do the delayed send for completely self-serving reasons. I don't care if anybody else is going to read it and get stressed out, because that's on them. They have the tools. It is on me. You made your choices. Now live with them. I live with them every day. OK, go ahead, Andrew.
Switching to Apple: Jake's Experience
00:02:16
Speaker
No, no. Speaking of focus modes, we're all on Apple devices, right? So Jake, you're the most recent addition to the Apple ecosystem. Have you done the whole setting up a million focuses thing?
00:02:26
Speaker
Yeah. So, uh, for those who don't know, I switched to Apple like two months ago after the being on Android for like five years after switching from Apple originally. Um, Jake, why did you switch? Which one, which way? Why did you switch to Apple? The first time I, okay. So I switched to Android because my wife's phone was stolen in Barcelona.
00:02:52
Speaker
And it was a whole like, uh, we have to get a phone, but there's a new iPhone in like a month and, uh, it doesn't make sense to get it then. Uh, I switched to Apple because I was sick of getting cyber bullied by my friends in group chats with all these tap backs, which I don't know if you know, if you have an iPhone and you thumbs up a text message from an Android user, they receive a text message, uh, for that thumbs up or they used to, and they're Google's trying to like get revenge about that.
Communication Platforms: Facebook Messenger Issues
00:03:22
Speaker
But I was so annoyed with it that I was just like, okay, fine. I'll get on this. My wife's on the iPhone and she can stop using Facebook Messenger to message me because she's so annoyed by my text messages. Oh my God. Facebook Messenger to message your wife. Yes. That's no way to live. Yeah, it was a bad sign, huh? Yeah. It's not home. Zuck is looking at all your dirty DMs that you two are exchanging.
00:03:50
Speaker
It's, it's going to be somebody, somebody's looking at every, I just assume right now that somebody's looking at everything that I write, which would, which is kind of a hopeful feeling also, you know, hopefully somebody is perceiving when I write. Just wait until your son figures out how to unlock your devices and go back and read through your text messages and to all that, uh, saucy mommy, daddy texting.
00:04:16
Speaker
Okay. I believe you. That's a good point. He'll see it when he'll believe it when it happens. Um, so we have like a couple of technology topics actually
AI in Legal Practice: Chat GPT's Role
00:04:25
Speaker
to talk about today. The main one we want to talk about, I think was AI. And since we're all three lawyers specifically how AI can apply to the law, I think, but generally just like messing around with chat, GPT and, and, uh, whatnot, and how in the last like couple of months it has become a useful thing that like normal people can actually use for something like in their work.
00:04:45
Speaker
Have you guys really gone crazy the last couple of months? Like, Dolly Mini was Dolly Mini. Did that break out a couple of months ago? That is for the image generator. Is that when that got really famous or was that earlier this year?
00:05:01
Speaker
Either way is this year. Yeah, it hasn't been that long. I mean, six months maybe at the most. I think the big thing that broke out was the stable diffusion, being able to do it on your device, on your own device, not having to have it hosted somewhere. That was the thing that sort of was a big deal maybe three or four months ago.
00:05:17
Speaker
Yeah. And so stable diffusion came out and DALIA I think is one iteration of the same sort of idea, but there are these AI image generation things. And now even more recently, like in the past, I don't know, three, four weeks, it seems like all of my Mastodon timeline, before that all of my Twitter timeline was talking about
00:05:42
Speaker
chat GPT, which I have no idea what chat GPT stands for. I'm sure one of you two knows. Andrew, you've got it now. What's GPT? Oh, I'm not Googling. Global partition table. Good protocol timing. Yeah. Okay. So I wish I knew. I don't know.
00:05:59
Speaker
Okay. Well, uh, whatever it is, chat GPT is something that has come out in more than just a casual social way, although it initiated for me in a casual social way because Andrew, I kept on seeing you, uh, typing out simple, stupid little things, uh, in, uh, the Slack that we're in together.
00:06:20
Speaker
That was like we have this topic of conversation and you would just disappear for 30 seconds and tell chat GPT to write an outline for this thing or write a paragraph about this thing. I guess tell us how you got started using chat GPT.
00:06:39
Speaker
Sure, so I found out what it stands for. It is generative pre-trained transformer three. So that was definitely on the tip of my tongue. Oh, that sucks. That's not nearly as cool as I hoped it would be. No, no, there's nothing in there. Well, transformer is kind of cool. That's sort of like a robot.
00:06:55
Speaker
I started using it. I got access to the beta for OpenAI, the initial GPT-2, or maybe one. But one of the earlier iterations of this type of model, this language generating model, because I just emailed them and said that I wanted to write an article about it. And I wound up publishing it in tax notes. And I was
00:07:15
Speaker
Part of that, you may have read the articles that were coming out maybe six months or a year ago where people would get an AI to write half the article. And then the headline after that would say, guess what? Did you believe everything that was just said above? That was all written by an AI or whatever. I did one of those sort of hackneyed articles for tax notes. Are you proud of yourself? I'm very proud of myself, yeah. I stand for it. Oh, I forgot about that article. Yeah, that's right. I had forgotten that you wrote that.
00:07:45
Speaker
Yeah, and it was good. It did a good job. It was plausible. It wasn't like a smarter child on aim or any of those early AIs that were just horrendous. You know what I mean? That would break if you just answered your name as a number. It just would never respond again. It would die. That was the first article that I remember reading that actually spooked me a little bit about AI.
00:08:08
Speaker
uh you know like yes yeah so good job i thought i was late to the party i thought everybody had already done that so because it wasn't like it wasn't an original idea of mine i had read i think like a salon.com article or slate or something where someone had like maybe a book review that half of it was written by the ai and then they you know
00:08:26
Speaker
you know, at the second half tell you that this is the prompt they put in and everything. But basically, the way these things work is they have huge language models that have just been like digesting piece of writing on the internet, and oftentimes, you know, copy written works, and not giving anything to the authors for for doing this, right. And it
00:08:46
Speaker
it can't think per se, but what it can do is predict how an idea would be conveyed in language. And so what chat GPT, all it really does is it puts it into a chat format. So you can just like, it's just like one of those automated little bots when you go to like a car dealership website or something that you can set schedules and, or I'm sorry, set appointments and things. It's just in that format, but it is interfacing with this huge model of generative language.
00:09:15
Speaker
And so the way I think I first saw it was people were using it
00:09:21
Speaker
to write stupid versions of Mambo No. 5. Write me a lyrics to Mambo No. 5, but have it be about librarians or something. And that's where I saw screenshots on Twitter. Around the time of the Twitter migration to Macedon, this was coming out around the same time. And so I think the idea that we want to talk about is how we see it being used in the law generally.
Chat GPT and Legal Writing: Trials and Errors
00:09:48
Speaker
You guys tell me, what are your thoughts on it?
00:09:50
Speaker
I started a little experiment. It was this week, maybe it was last week where I had a brief due summary judgment brief due on a case and I thought what better time to start trying out this particular brand of nonsense.
00:10:10
Speaker
And so I started feeding into it this stuff about my brief to get it to generate paragraphs, to get it to generate like an introduction. And an introduction was about all that it was useful. So it was, you know, a story about how this employee had experienced sexual harassment in the workplace and hostility on the basis of sex with comments about like hand size and all the innuendo and nuance that goes along with that.
00:10:39
Speaker
starting to drop in there the names of the particular people and pretty quickly bumped into the limitations of the program where it basically said something along the lines of like, I'm sorry, I can't write a work of fiction for you or something like that. I bumped into the walls pretty quickly. What I did manage to get out of it was a pretty serviceable introductory paragraph or introductory section where I think the prompt I fed to it was,
00:11:09
Speaker
Give me 1500 words on the prima facie case for proving sexual harassment or the prima facie case for establishing a
00:11:24
Speaker
an initial case for a sexual harassment claim. And it did that pretty serviceably. You had to go back through and pop in the right names for people. You had to go pop in specific details because it's not really good at handling the details or specifics of a case that you feed into it. But it wrote a serviceable introductory paragraph that ended up getting rewritten a couple of times.
00:11:51
Speaker
useful, maybe a little bit. Uh, it's not writing briefs for me. That's for sure. At this point, it definitely feels like a proof of concept though. Like I think the thing that's really made chat GPT take off is the fact that it's really pretty good at sounding good.
00:12:10
Speaker
at and doing kind of what you want at least from what I've seen I personally haven't used it but I've seen you know stuff that you made it right Andrew and then I saw other people asking it to write stuff about like corporate law and it would produce a summary that was maybe a little off but I don't know about you guys but I spend a lot of time writing emails just like basic stuff
00:12:34
Speaker
uh where like i'm just trying to write something a summary that doesn't sound stupid um and this might be able to like something like this might be able to i don't know maybe take some time i don't in my practice it doesn't really make sense
00:12:48
Speaker
It's definitely not bulletproof. Yeah, I know. And you would definitely have to like go in and double check anything, like both for like ethical reasons, but also because, you know, who knows what's pulling from the ether. Yeah, I did another experiment with it where I just invited it to and this was just completely fooling around, not related to anything.
00:13:11
Speaker
But I took and gave it a prompt, something along the lines of write a thousand words about the state of landlord, residential landlord tenant law in the state of Indiana. And it got a lot of things right. It figured out the security deposit handling and that you have to write a letter. The landlord has to write a letter within 45 days detailing what the damages were for the property.
00:13:37
Speaker
then if they don't return the security deposit or provide that accounting of what happened with it, it got a lot of stuff right in an impressive way. And I shared it with someone who does a lot more landlord tenant stuff in Indiana than I do. And he was like, it's not 100% incorrect, but there are two really big problems here that I can point out in seconds. So it's not a lawyer. It's not practicing law for you yet.
00:14:02
Speaker
But if you need it to churn out like a lame thousand word blog post, uh, you can start with this and go and fix all of the stupid stuff that it gets way wrong.
AI's Impact on Content Writing Jobs
00:14:14
Speaker
Yeah. Maybe this eliminates the like buzzfeed, uh, or not. Buzzfeed might be the wrong example to use like those content farms that pay like, you know,
00:14:27
Speaker
pittances for people to just pump out words to reduce their SEO or whatever. I did that in college. It wasn't like it wasn't really a pittance at the time. I mean, it was I think it was twenty five dollars an article for like five hundred words. And I could bang out, you know, twenty five or thirty of them in a day. So it wasn't bad. And they were. But they were all like insane props. I mean, like how to get ground beef out of, you know, carpet or something. And it's you just got to come up with some. You know what I mean? Some some plausible sounding ground beef out of carpet.
00:14:56
Speaker
I think I talked about using vinegar or something. Maybe get a dog. I don't know. There's like your enthusiasm episode about this, I think. Is there? What I've heard it likened to the chat GPT is like mansplaining as a service because one huge problem with it is it, uh, it doesn't give any indication as to how certain it is of the answer it gives you.
00:15:19
Speaker
So it says it authoritatively and clearly as though this is definitely the answer. And there's no indication that it's hedging on this or it's not certain. And I think that would be a huge addition just to sort of like, even if it was like a color coded thing where there's sentences that check the factual accuracy of this because I'm not certain.
00:15:39
Speaker
Yeah, I think that would add a lot. It would be great. Anyway, go ahead. That's something that cars do. If you're in a Tesla vehicle and you've got the full self-driving mode active, full self-driving. I'm doing air quotes here because it's not. But if you do the full self-driving mode, it'll show you different color codes on the road that it's visualizing for you and then spitting up onto the screen.
00:16:02
Speaker
if it's in green or I don't all have the colors wrong, but if it's in green, like I'm absolutely certain that this is where the curb is. And if it's yellow, it's not quite certain. And if it's red, it's like, there might be a curb here. And so to even have some sort of visual cue that's available to the user about here's our degree of certainty about the correctness of the content that we put into this thing that you're using here, like
00:16:27
Speaker
I have to imagine that that exists, that there's a degree of certainty that's behind the scenes that's not exposed to the user, but that would be an interesting addition for, I don't know, chat GPT power users, maybe? Who knows?
00:16:42
Speaker
doing a legal brief so it might not even understand the concept that it should even question how sure it is about a thing because it's just writing as far as that's concerned. It's not different. It's the same as tell me a story. So there is no certainty involved.
00:17:01
Speaker
Um, I, one thing that I was thinking about is I was, we were told in law school that we should be worried about automated e-discovery, that it was going to take over, take a bunch of lawyers jobs. And that hasn't happened as far as I can tell at all. Um,
00:17:16
Speaker
But and chat GPT is like more flashy because it's a creative output.
AI in Legal Discovery: Future Possibilities
00:17:21
Speaker
The output is creative as opposed to, you know, algorithmically looking over documents. And I don't even know how AI based discovery eDiscovery would work.
00:17:32
Speaker
That's what I was going to ask you. You said it, and I understand I should be afraid of it. I agree, I guess. But I don't know what you're talking about. How would that work exactly? Just crawling over a server farm of data and pulling out the relevant parts? Well, I guess to some extent, it might already happen. So you get a packet of 2,000 pages in PDFs right now. The first thing I do is I run it through a text recognition.
00:18:00
Speaker
scan if it's not already if it's not already RCR. And then I search for specific keywords I really know I need to look for. And sometimes that's enough to at least get me started on something. And that takes
00:18:15
Speaker
considerably less time than what they used to do, which is get giant boxes and have associates pour over them one by one gradually. So in that way, it's already the OCR process was automated, but really, it's barely not automated at all. So you take that one step further, you get PDFs, you say, run these through OCR, here are my keywords, give me the responsive pages.
00:18:43
Speaker
And then that's some amount of time save. But it could go even further and be like, here are the keywords in the complaint or something, I don't know.
00:18:55
Speaker
but I think you'll still be checking over the work. I mean, so I've done a little, I've done a little bit of like sort of experimenting. I dumped in the, uh, the, the regulation, um, for what constitutes a business expense into this chat GPT. I like dumped the whole, like literally control C control V and then, uh, ask the question at the bottom based on this. Can I, if I'm a glassblower, can I take a, um, business expense for a laptop that I won't use for my job or something? It's something like pretty straightforward and it,
00:19:24
Speaker
it answered properly based on the content I fed to it. And so I can see how there could be some utility there. And I've done things like I've messed around a little bit with, I dumped my resume into it and said, write me like an introductory executive summary for the top that says like, you know, what the hell I've done or whatever.
00:19:41
Speaker
it's pretty good at taking like you feed it information in one format, and it takes it and sort of just literally paste it into another format. But I don't know, there may be tools out there that can do that, that like an AI may be like using a bazooka to kill a fly with that. Like I don't know if you need like this huge language model to do something like that. It's basically like a template or a glorified macro or something. Yeah, so that is one really interesting and useful area where this is developing right now.
00:20:10
Speaker
Uh, we call it AI. I don't think any of this is actually AI and we could get into a huge semantic thing about it, but everybody's calling it AI. So we're going to call it AI. It's probably more like machine learning, but AI is more buzzword. A complex algorithm, like this one math equation.
00:20:27
Speaker
But I bumped into and did a little trial of it. It didn't work probably for clumsy onboarding reasons, not because the software is bad. But I tried this software that basically what you do is you're writing a brief. I was writing one in October or November.
00:20:46
Speaker
and you upload all of your information from the case. You upload all of your deposition transcripts, all of the evidence that's been tendered to you, all of the interrogatory responses, and then you write your brief and then you run this software over it. And supposedly what it's going to do is go through and cite through all of that material that you've uploaded and it's digested
00:21:09
Speaker
and then put pinpoint sites in this deposition of whatever person at this page, these lines, this is the factual support, the evidence that you want to designate as factual support in this motion for summary judgment.
00:21:26
Speaker
opposition to a motion for summary judgment. It'll do the same with case law where if you're making a point and you just put in some placeholder text like site to legal authority here and you've uploaded the PDF as part of your packet, it will theoretically pop it in there.
00:21:43
Speaker
And so that's the kind of stuff that I'm really excited about where, you know, if I'm spending between 20 and 50 hours working on emotion for summary judgment, something like half that time, maybe more isn't writing. It's pouring over those deposition transcripts, flagging the stuff that I know is in there and I can write in the brief, but I actually have to spend the hours, you know, command F command V or command F command C command V, like,
00:22:10
Speaker
copying and pasting that stuff into the brief. And so that's one of the uses of AI where not generative like chat GPT, but referential where I can give it a body of work and evidence to feed into it and say, I'm going to write this thing. Now you go and do all of the work of citing to the evidence. I'm really looking forward to that getting really, uh, um, narrowed down and like on target. That's going to be great.
00:22:40
Speaker
You know how I know you don't use billable hours? You're excited about all this and this isn't like the scariest thing in the world. I say that, but like all this busy work that he would take care of is like everybody's least favorite thing to do. But like, yeah, it's mostly for a contingent fee attorney, it's gotta be just a total value add to not have to do all this stuff.
00:23:06
Speaker
uh all this busy work that isn't the fun part of lawyering the fun part is the thinking and the puzzle solving and the you know argument or whatever yeah uh copying and pasting sites and making you know double checking your blue booking and so it only only freaks love that part uh i'm talking about myself a little bit but not not all the time
00:23:28
Speaker
Oh, yeah, I definitely don't relish the going through the brief that I've written and checking to make sure that every single one of my citations is in the designated evidence and every single, you know, everything is in there is actually uploaded to the court eventually, like, if I could do away with that part of it.
00:23:47
Speaker
God bless them if they can figure it out. I'm not going to name drop them on here because I don't want to be overly critical of them because it wasn't quite ready for actual real life use. After I tried it, I decided not to pay for it yet. I'm excited about the prospect though. They're not zeroed in, but they're getting close and I'm excited for when that day arrives.
00:24:11
Speaker
I was thinking that the true customers for these or the first customers for these might be like the huge volume firms like the Morgan & Morgan's of the world who have like thousands of lawsuits going at any given time. But they probably already have a pretty good system of automation even without the algorithms.
00:24:35
Speaker
Like they are guaranteed to have a million motions for summary judgment on any all kinds of law. Cut you ready to copy and paste. I'm guessing I, you know, some of my best friends worked for Morgan and Morgan. But I don't know defensive. But I don't know. I'm sure that they have some kind of efficiency there just because for the sheer amount that they do.
00:25:01
Speaker
Um, and copy and paste is, I mean, better than an AI new generation a lot of the time. So you just got to plug it by your, your cases and your site in your, in your parties, et cetera. In 15 years, somebody's going to come back and dig this old podcast episode up and be like, look at these old guys who are so skeptical of AI. What a bunch of dinosaurs fools. Yeah.
00:25:26
Speaker
In my field, we still have people who are only reachable by fax machine. You mentioned Andrew, you thought you were behind on writing about AI. I guarantee you if I asked the majority of my clients, they would never have heard of chat GPT still. We are terminally online, terminally trying to stay with it. But majority of people are not hearing about this.
00:25:56
Speaker
Yeah, I mean, I think this was pretty mainstream and that is only mainstream online, right? Because I mean, I somewhat regularly work with older tax attorneys that have, they're at larger firms and they don't write their own email. Like I don't necessarily find this out for some time, but one in particular I'm thinking about, he doesn't write his own emails. They get printed out, he reads them, he writes something down on his paper and he gives it to his secretary. And so I'm not really communicating, because like I was having some sort of back and forth with him at some point.
00:26:25
Speaker
And she like broke the fourth wall and said like, um, I, you know, I'm here, but I can't reach him. He's not, I'm sorry. And I was like, I'm sorry. Who, who is the, well, he doesn't, he doesn't write his own emails. He doesn't read them. He, I print them out. Pay no attention to the man behind the curtain.
00:26:40
Speaker
Right. He could have been dead for 20 years and this person is just keeping this whole practice going. Who knows? I'm just the wizard. It's the man behind the curtain and the man behind the curtain. He's not responding right now. Right. It's all paralegals all the way down. He's asleep. Yeah. I think the real value for the chat GPT, to your point with Morgan and Morgan and these other types of things, what I see is
00:27:04
Speaker
If you could expose it to your own data, that's going to be the interesting thing.
Personal Data and Productivity with AI
00:27:09
Speaker
If there was something like stable diffusion where you could run it locally and you could just basically use it for like even just searching your own system for emails, for turns of phrase that you use, for like even just as glorified search, it would be fantastic. If you ran chat GPT on your own writing, would it be like when you listen to your own voice on a recording and you just hate it and you're just like, oh, this is the worst thing I've ever heard?
00:27:32
Speaker
It might be, yeah. Or it validates you, and it's like, oh, this writing's so good. Oh my gosh. It's like that we talk about the YouTube algorithm radicalizing you. This would radicalize you with your own writing. It would just keep feeding your own opinions back to you, slightly tweaked up, and you would get more and more incensed, and it would just, yeah.
00:27:53
Speaker
Yeah, this is a literal echo chamber like literally your own voice bouncing back over and over Slightly different words and that's it. Yeah. Yeah Sell me on that open AI. Oh I so open open AI is this is costing them money right because it's all free right now. I saw some estimates that Chat GPT is cause it costs like three million dollars a month to run or something like that
00:28:21
Speaker
Which it doesn't seem like that much. Well, I mean, I think they had some, they were having some big traffic problems there for a while. You would try to use it and it would say, I'm sorry, we're under heavy, you know, load and stuff. So I think they've been basically adding servers pretty regularly. And yeah, I think they're just burning through, um, VC money because the problem, I mean, so the, the joke answer I heard, I don't know, this may be apocryphal, but I heard someone say that the CEO was quoted in an interview.
00:28:47
Speaker
This is heard someone say, heard someone say. So if he didn't say this, then he didn't say this. When it comes time to monetize, they would ask the AI how best to monetize the system. And so he may have been a little, you know, being cheeky about it. That's clever if he actually said that.
00:29:03
Speaker
Yeah, I don't I do wonder how they're going to like my what I guess is initially they'll charge for API access So maybe they'll keep chat GPT that interface like the chat interface open but if you want to say streamline it into Microsoft Word so that as I begin to type, you know deer so-and-so or whatever it
00:29:25
Speaker
sort of predicts my letter for me. You know what I mean? It rolls in the AI language generation stuff into some other program you would need to pay for the API access. I think that's a way that those sorts of services often go. And that is how they've monetized the earlier GPT models. There's several other models that have not been sort of forced into this chat system. And so they're a little bit more kludgy to actually use.
00:29:52
Speaker
They have so many API calls or so many calls in general per Month that are free and then beyond that you have to pay and so there are multi-million dollar startups that have you know multi-million dollars VC funds So like who knows this is really multi-million dollars barely exists. Yeah monopoly money multi-million dollar startups that have cropped up and sold already one of them that comes to mind was a copy generation of
00:30:17
Speaker
software as a service. Basically, you enter some details about your company. It generates all the copy you need for the website, and there you go. And so that company was making, I don't know, $100,000 a month or something. And they sold to one of the big, like, Andreessen Horowitz or something. And so I could see something like that, API calls. But yeah, as of right now, they are unquestionably burning through millions of dollars, some of which is, I believe, Elon Musk's money.
00:30:46
Speaker
Oh, great. OpenAI, he was involved in that somehow, right? Yeah, he's got some weird thoughts about AI. He sort of thinks Terminator is a documentary and he's afraid that things could actually go that way, but he seems to be wanting to fund it also and he's made some... What's he doing with OpenAI then?
00:31:07
Speaker
Chat GPT is going to throw out the robot manifesto at some point. Then we're going to have a real problem on our end. It's over. Yeah. Our chats will never be the same. Yeah. So one of the best implementations of kind of procedurally generated stuff that I've seen is pops up on the esq.social Mastodon instance recently. And David Colarusso, who is just a brilliant dude and like, I,
00:31:37
Speaker
Sometimes you see people online and you just think, I want to be this guy's friend. Like this guy seems great. Uh, David call it Russo is one of those guys for me. I would totally love to be that guy's friend, but he has generated on the pod. Let's get him here. Yeah, sure. Sure. We're calling it. We're calling it a pod. We're crossing that line, huh? 31 minutes in and we're a pod. Yeah. Yeah. We're a pod.
00:31:58
Speaker
But he has created this ICYMI, in case you missed it, law bot that basically has taken and distilled down to an account what would have been a much healthier version of like a social media algorithm. And so it crawls through the posts that have been made on a particular mastodon instance for a day, things that have been posted, things that have been boosted, things that have been, I don't know if it covers things that have been favorited.
00:32:28
Speaker
Uh, and it digests it and distills it down to basically the best few topics of the day and kind of re boosts those things. And then it also procedurally generates it into a five minute.
00:32:45
Speaker
podcast, five to seven minute podcast that you can just listen to that will verbally cover the best news of the day according to this bot. I think that he's going to learn from it, he's going to improve it, and there's going to be iteration on it.
00:33:03
Speaker
But this is, I think, one of the best applications of kind of machine learning to social media, where he created this bot that just does this great thing of distilling a bunch of things that a bunch of nerdy lawyers are saying into a verbal medium.
00:33:20
Speaker
I think that's really cool. I agree. I said this on the on s.social or mastodon instance. But I told him like, I think that this is it's really filling the void on mastodon, which is mastodon doesn't serve you anything.
Social Media and Personalized Feeds: Mastodon
00:33:37
Speaker
It's not pushing content to you. Right. And as much as I hated Twitter, like constantly trying to show me content about Connery's pants destroying a Minecraft castle or something, something I didn't care about at all. I did miss seeing stuff that I wouldn't have seen otherwise. And
00:33:57
Speaker
Like his bot really taught me, oh, there's a world where you can opt into algorithms. You can be like, no, I don't want to see whatever political thing. I don't I don't want to see Minecraft influencers fighting. I only want to see my influencers fighting. I only want to see, you know, in case you missed it, lost stuff. And I am really hopeful. I'm sure he's not the only guy around doing this kind of thing.
00:34:27
Speaker
But man, at some point I can just follow a bunch of bots and that'll be my algorithm content. I can even have it as a list and just have an algorithm feed and my people that I'm actually following on purpose feed and have those basically do everything that Twitter wants to do. Everything that the algorithm tries to do for you. Did you just come up
00:34:53
Speaker
Did you just come up with that phrase yourself, the opt-in algorithm? Because I love that. I think that's great. Yeah. I said it to him earlier, like two, like three days ago or whatever. But yeah, if you didn't just coin that phrase, if you didn't just coin that phrase, you absolutely need to like trademark it right now. Unfortunately, Kola Russo already knows it. So he's gonna, he, he's a smart enough guy to go grab that. So.
00:35:17
Speaker
You know what? He deserves it. He does. He deserves it. Yes. I don't know. You mentioned that if you're sure other people are doing it, I have not seen any other similar bots. There may be other people doing it. I'm sure people are thinking about it or working on it, but he's definitely as of now on the bleeding edge, I think. And yeah, I couldn't say it better than you said it, Jake.
00:35:36
Speaker
It the idea like the the aspect of twitter that surfaced things that you had missed was not the problem To me anyway, it was that it was who was deciding what it was that I had missed And how it was often not even anyone I was following right? This is something where it's like it's limited to just these what's one instance or he may have actually exposed it to Like legal that social or or law builders something or other so there may be a couple different instances But it's like a closed universe of what is going what I what I missed
00:36:06
Speaker
Right. Also, he follows specific accounts, I think. So he exposes it to both instances and accounts. And so if there's any like spit, like you could have like rock band bot that follows only Paramore and, you know, whatever, whatever other rock bands don't exist.
00:36:25
Speaker
Of course, Paramore is your example. Yeah, of course. Paramore and Radiohead, the only two important rock bands still in existence. And you're the youth correspondent. Yes, I'm the youth correspondent. The youth don't listen to rock. I don't know if you know this. The youth listen to something called trap.
00:36:47
Speaker
Trap. Trap, yes. I don't know what that is. That's one of those things that I've heard it for like 15 years be referred to. And I should have asked that meme of like, it's too late now to ask. I should have asked 15 years ago and now I just have to kind of nod along. Oh yeah, sure. It sounds all sounds like trance music or like techno to me. I can't tell you. No, no, it's not even close. It's not even close. Yeah, man. How is it that I'm the oldest person on here and I am the most in touch with the youth? That's not true. We need to revisit.
00:37:16
Speaker
We need to revisit your youth correspondence status here because it's not it doesn't seem to be working out I'm a mere 35 years old. So I get to say I'm the youth correspondent Okay, that's how it works
00:37:33
Speaker
So on the topic of AI, one sort of issue that some folks are concerned about, and I was wondering what you guys think about it, is the concept of AI is basically stealing other people's
AI Art Generation: Innovation or Theft?
00:37:45
Speaker
work, right? And so there are two, not I won't say applications, but two sort of fields or areas where this is most apparent, and I hear it the most.
00:37:53
Speaker
One is in art in like generating like like graphics like visual art and the other is in code because I've also experimented with a thing called github copilot which I'll be really quick with because it's geeky and boring but basically what it is is it's a plug-in for
00:38:11
Speaker
Microsoft visual studio code which is a an application for writing code and what copilot does is if you just write in comments the description of what you want a particular like module or sub portion of a program to do it will sort of auto generate what it thinks is a pretty good approximation of.
00:38:31
Speaker
how you would go about doing that. And it really is pretty useful. It's good for checking your code. It's good for getting some ideas, because a lot of times half the deal, half the sort of work in writing a program is figuring out how to do something. Like you know generally I have to write a module here that will hash this input of a password into just its hash format. It's sha256 or something. And you know that that's what you want to do, but you don't really know how to do it. And so long story short, it can do that for you.
00:38:59
Speaker
It's pretty good. The big complaint, and there's a lawsuit pending for it, is that basically every GitHub repo and GitHub, for people who don't know, is just basically a giant repository of source code. People who are willing to open the source code to their programs and show people the recipes for making these programs just upload it to GitHub, and they're all in these repositories. GitHub Copilot was created by exposing an AI to all of that code. And so the concern is if I
00:39:27
Speaker
have the AI generate a login screen from GitHub co-pilot, it may just be pulling the login screen that you meticulously wrote for your program that is uploaded on GitHub. It's not necessarily even homogenizing multiple login screens the way
00:39:45
Speaker
one would assume chat GPT is, right? Like everything chat GPT is saying to you, it has seen somewhere else, but you hope that it's atomized sufficiently that it's not taking whole sentences from copy written works. With code, there's only so many ways to do a lot of different things that you have a much bigger chance of running into. Like literally, this is a line of code that you took from my GitHub repository, and I didn't give you permission to do that.
00:40:10
Speaker
And the same thing kind of goes for art. And so the general idea, the general topic I thought we could talk about, or it wasn't actually my topic, it was either the AIs or one of you guys, is just the idea, is AI stealing work from people? How do you sort of square that circle to say that that's okay, right? I think I put this on here. And by the way,
00:40:33
Speaker
The code that people put onto GitHub, they get to choose the license that other people get to use it under, right? So they can be like, you only get to use it if blah, blah, blah. And then so commercial has to be attributed.
00:40:48
Speaker
Yeah and then co-pilot might steal that language and it doesn't give you any of that attribution or anything like that. The license might be violated even if it's not a cut like I guess that would still make it a copyright violation because you're violating the license.
00:41:07
Speaker
Uh, but yeah, that's another concern is that it's taking all this language without regard to the license and possibly we I guess well, we don't know Uh filtering it through to other to people who want it. Um, yeah, I think their argument is that it was only you know, uh, just to be clear I I think that that's sort of the the crux of the lawsuit I think their argument is that it was only things that were issued under the public domain and so like they could do this they could you know more
00:41:33
Speaker
or one of the Creative Commons licenses that allows no attribution and commercial use or something. I believe that's what they're saying and I think that's sort of where the argument is. But the other side, the concern is that I think some people found some lines of code that seem to come from things that were non-commercial
00:41:52
Speaker
attribution required and obviously there's no attribution because it's just plucking that line of code. But yes, I'm sorry, go on. Yeah, I put this on the list because it reminded me of...
00:42:05
Speaker
It reminded me of the opinion of Judge Kaczynski, which you may remember from where it was Vanna White sued somebody over an ad which had a robot version of Vanna White doing a Wheel of Fortune kind of thing, saying that it violated her right of publicity or something. It was clearly not Vanna White, but it was obviously a reference to her.
00:42:32
Speaker
And Kaczynski actually wrote it a descent talking about how important it is that people be able to, that there be a healthy public domain of works that take inspiration from copyrighted work. And so if you were a human and specifically talking about an AI creating art, if you were a human and you were looking at like a modern art thing,
00:42:56
Speaker
You could and it's like, you know, just a bunch of solid reds and some oranges in a pattern. You should be able to take inspiration from that and do something very similar. And I think the concern here is that so this this isn't that different from it if it's actually transforming it. It's a little different if they're just like stealing lines of code from right code repository. But like art is like looking at the general
00:43:26
Speaker
artistic configuration of something and then transforming it into something that looks like it. And it's kind of like putting it through a filter. It's similar to what people do with inspiration in public domain. I think the main worry that a lot of people actually have, the thing that's like cutting to the core, is that a lot of people don't really care about the artistic process.
00:43:49
Speaker
And it might become so the the ease with which you can create art with this might become like a real problem.
00:43:58
Speaker
And that is a little concerning to me. I don't know what the legal response to that could be. The idea that it just becomes so easy to make art that artists get put out of work and then the universe that is art gets smaller because fewer people are doing it. It's not great.
00:44:20
Speaker
That's a good point. Yeah, go ahead. I think it has already eaten the lunch of a lot of stock image services and things like where you don't particularly care. You just want a picture of a man standing in a shield. A picture of a man standing in a field. If you're eating Getty Images lunch like Godspeed, go for it.
00:44:43
Speaker
I agree. And I mean, I think sort of to Jake's point, it's not as though those stock image services were cutting huge checks to these artists that were the original photographers or painters or whatever the hell it was, right? So it's not like it's really taking money out of artists. That wasn't taking... Like shutting down stock image companies is not going to be taking very much out of the pockets of individual creators anyway.
00:45:06
Speaker
Yeah. Yeah. And I also all this A.R. A.R. for now, it looks pretty similar to each other. And so I can't I can't totally tell or I can't tell with 100 percent certainty when something's created by a. But I have a pretty good idea because there's something a little weird about it a lot of the time. It's like the uncanny valley. Yeah. There was this fantastic Twitter thread.
00:45:34
Speaker
There was this fantastic Twitter thread about Kermit the Frog. Like show me, create a picture of Kermit the Frog doing this or that in the style of this or that. And it was like, show me Kermit the Frog in the style of like, I don't know who's the famous anime director who does the Spirited Away and all that stuff.
00:45:58
Speaker
Yeah show me Kermit the Frog sipping Lipton tea in the style of Miyazaki or something like that those get really really good and they honor the description that you're giving to it but you know I may be the dork on here who's gonna quote the Bible but
00:46:17
Speaker
I can't help but thinking of Ecclesiastes where it says, what has been is what will be. What has been done is what will be done. There is nothing new under the sun. And that's what this is about. So this is taking old things and splicing them, combining them, iterating on them in a way that's new and interesting to some people.
00:46:42
Speaker
Is that a copyright violation i don't know like is it a copyright violation when i go and i look at i get this idea somebody prompts me somebody gives me a commission to draw a picture of kirmit the frog in sipping lipton t in the style of me as a key like. Obviously i've been given all of the parameters here i'm supposed to mimic this art style and mimic this character am i committing a copyright copyright violation.
00:47:11
Speaker
maybe, I don't know, is Kermit copyrighted, but I'm not a copyright lawyer. I see people's bad voices. None of us are IP lawyers. None of us are IP lawyers. Yeah, this is definitely not copyright legal advice, but there's nothing new under the sun. The AI is doing something that a human could do and would do if somebody paid him 50 bucks on DeviantArt or whatever other
00:47:35
Speaker
Etsy, is it better or worse for that to be done by an AI or a human?
00:47:43
Speaker
I don't have the answer to that, but there's nothing new. If a human did it, I would say that's really cool. If I saw a human that a human work of Kermit the Frog in the style of Miyazaki, I would have been like, wow, great job. That's super cool. I love this. I thought you were going to talk about the phenomenon, how certain art algorithms, if you say the right words, they all reproduce a very spooky woman.
00:48:08
Speaker
I don't know if you've seen that thing. I have, yeah. That is very freaky. It's like some mathematical average of a woman's face, and that apparently is very disquieting to the human psyche. Is that the thing that was circulating months or years ago, like Momo? Is that what that is? Yes. Oh, yeah, that's horrifying. I think it's Momo. I'm a no-no for Momo. I've been trying to get a... It's close.
00:48:38
Speaker
I've been trying to get Crayon to make Momo and Crayon being what they call Dolly Mini now. And I tried a little bit. But yeah, maybe they have anti Momo technology now. I don't know. I mean, I think to your point, it's like with nothing new under the sun. It's a tool like any other, right? And I remember when Photoshop
00:48:59
Speaker
was first beginning to really get good. There was the same discussions of you're not going to be able to trust anything you see anymore. That's it. You know what I mean? Every photo you see is going to be touched up and airbrushed and everything. And to an extent, that's true. And we have seen that. But it's an interesting question as to what can be done about it. Is this an unalloyed good? Maybe not. But it's happening, right?
00:49:26
Speaker
What are we really going to do about it? And I don't think there's putting there's any way to sort of like put the toothpaste back on the tube on this. I mean, these AI generated art, the art generating things, they've already ingested all of the the collected world knowledge and every piece of art that is available online. It's over. And so I guess there will be lawsuits, right? And I'm sure there'll be some money exchanging hands and things, but it's not going to be undone. They're just going to pay for the right to continue to do this.
00:49:52
Speaker
So I don't know a lot about copyright, but I do know one phrase that is, I promise, taken from copyright law is slavish copy. And the real question to me, it seems like as somebody who doesn't know the first thing about copyright law, except for a very limited thing. So maybe the first thing, but not much beyond that.
00:50:12
Speaker
If it is a slavish copy, like taking that one line of code from somebody's GitHub commit and just duplicating that, that's a slavish copy. But if somebody is, or if a person or a machine is iterating on something in a meaningful way that changes it and adds something to it or comments on it in some way,
00:50:37
Speaker
I think that's interesting and at least worth having around and probably not a slavish copy in the sense of literally all you did was make a Xerox of this thing. That's the kind of thing that we're worried about.
00:50:53
Speaker
probably bound to happen where you end up with those straight up copies. Um, and those ones probably do deserve to see some repercussions for just the straight up copy, but meaningful and interesting iterations. Like, uh, I just, I just had it pulled up here. Uh, a still of Kermit the frog in blade runner 2049, like
00:51:16
Speaker
That's cool and interesting and I want that stuff to exist and it can exist if a computer generates it or a human generates it and I'm excited to see it either way. That's cool stuff.
00:51:31
Speaker
I want to see that. It's not long. I think it's not long. There have been some applications that have like small animations, like very low frame rate, simple animations. But I don't know if you guys have seen, it doesn't, they don't, Dolly doesn't handle like human faces very well. And so I've seen some of the more famous examples. Like one of them was like,
00:51:51
Speaker
of show me Danny DeVito eating a banana. And like his face and the banana are not really discernible as being separate. It's a horror show. It's the Momo thing, right? So I think we're a ways off before those sorts of things can have, like we can have, you know, full animated human beings, photograph style, you know what I mean? Like photography quality type things. But yeah, I agree. I think this is, it's a tool like any other. It's
00:52:20
Speaker
I don't want to say it's much to do about nothing, but the hand wringing that seems to happen every time something new comes out as to, you know, this is going to put tons of people out of work. This is going to, you know, close doors for a lot of different people. I mean, the same thing is going on in the coding community. There are some people that are concerned that AI is going to take their job like simple coders jobs, right, because it can be done by
00:52:43
Speaker
GitHub co-pilot or honestly chat GPT. But I don't know if you guys have seen it all. The other thing is that there's a whole marketplace for prompts. People are selling just the prompts required to
00:52:57
Speaker
generate a certain piece of art because it is like its own thing, its own skill to kind of learn how to, you know, how to get the perfect Kermit drinking tea. And so the analogy I've heard made there is like, how is that like the initial, the first cut of that is like, that's insane. Who's paying for these prompts, right? But how is that different from source code for a given program? Really, right? This is the code that generates this program. This is the code that generates this image. What's the difference? I would love to see how long these prompts that are that that would generate like
00:53:27
Speaker
That's a good question. That would justify paying for it. But also, would it not just produce the same thing over and over? I guess I can tell there's some degree of randomness in Kryon. Yeah, it doesn't generate the same stuff. Yeah. There's definitely some degree. It arrives at it, I think, through different routes, basically. Like, mathematically, you know what I mean? Like, it takes your terms and goes in different directions with it between different times. Yeah.
00:53:56
Speaker
But we're running a little low on time here. But the
Facial Recognition and Legal Events
00:54:00
Speaker
other thing we really want to talk about, which I think segues right in, because it is basically technology run amok, is the story about the lawyer who we should sort of set it up a little bit in case people haven't read it. There was a lawyer from New Jersey, like me, New Jersey, hey, look at that, who went to, I think, a ruckus show.
00:54:20
Speaker
at Madison Square Garden. I think she was with her children. It's as bad as it could be in that regard. And she was pretty quickly beset upon by the security guards. And they said, you have to leave because your firm is engaged in some sort of litigation with, I assume, MSG, Madison Square Garden, or whatever the parent company of that is.
00:54:42
Speaker
And she was escorted out. She said that, like, she doesn't even, I don't think, practice that type of law. Someone in her firm, I think, is engaged in litigation with MSG, but she wasn't. And so this was sort of a minor kerfuffle in the legal community for a little while. What do you guys think?
00:54:58
Speaker
Yeah, well, it wasn't even lawsuit against Madison Square Garden. It was a lawsuit against a restaurant that MSG Entertainment owns in a different state with a different attorney, like totally unrelated. I think it was actually Radio City Music Hall. But yeah, she was totally, there is no logical connection between her and that lawsuit.
00:55:24
Speaker
But yeah, of course, the most the tech at hook with this being that they discovered her through facial recognition, which is kind of like, which I guess they started earlier this year, kind of the confession that they don't they don't actually have a good justification for this other than if you sue us, we're going to we're going to harass you. We're going to we're going to feed it. We're firms headshots and just let you go see the rockets.
00:55:52
Speaker
Yeah, no sane person does that for like good reasons. And they're getting sued, by the way. I don't know if you knew that, but no, this she is not the first attorney to have this happen to her. And they're getting sued by multiple law firms on multiple theories. One of them is saying that they should lose their liquor license because the liquor license has like public accommodation requirements or something like that, where you're allowed to kick people out for specific reasons.
00:56:18
Speaker
Um, and they're, they're alleging that, or they're, they're saying that it's like on the contract when you purchase a ticket, right? Or it's like on the back of it. It's like when you go to a baseball game, they say, if you get hit with a foul ball, it's that's on you or whatever.
00:56:29
Speaker
Is it? I don't know. I think that's what they're saying. Oh, okay. Yeah. They said that this is like a security concern or something like that. It's a naturally oppositional environment when the lawyer that's suing us comes in, which is like such BS. I don't know. So James Dolan runs MSG Entertainment. James Dolan.
00:56:55
Speaker
who owns the Knicks and owns the Rangers and is famously everybody hates him for many reasons not just for being a bad owner like in terms of producing wins but like is apparently just a big ass and that's just like his thing and so this is not this is completely consistent with his character
00:57:20
Speaker
Jake, I hope you never wanted to go see anything at Radio City Music Hall. Congratulations. You're on an exclusion list. Yeah, please. James Dolan, I'm coming for you. I'm not afraid of you, James Dolan. Oh, no. This is one of the reasons why some people like to get rich is so that they can do whatever they want to people and can bully
00:57:44
Speaker
And I know it's a tech story because it's kind of freaky that like, oh, dang, I'm on a law firm because my face exists next to my name. People can know where I am. And that's freaky a little bit. But I'm glad that we are kind of acknowledging it because, man, that information has been out there for a while, I feel like.
00:58:08
Speaker
A couple of other interesting tidbits in this to just really zero in, and we'll put the link to this article in the New York Times in the show notes here. She was chaperoning her nine-year-old daughter's Girl Scout troop on a trip to see this.
00:58:24
Speaker
And so I have a 10-year-old and an 8-year-old, so I roughly understand what the implications would be here. Huge embarrassment for the parent, huge embarrassment for the child. And what happened was the security guards identified her using a facial recognition system. And apparently, she was on something called an attorney exclusion list,
00:58:49
Speaker
which doesn't just bring in the people who have ticked off James Dolan. It brings off all of their law partners, all of their associates, anybody who's in a firm that is associated with a lawyer who has ticked off James Dolan. And this is in a spot where we're talking about Radio City Music Hall, which is one of the fabled concert venues in New York City and probably the world.
00:59:16
Speaker
And now somebody can be excluded from there using facial recognition software not because they've done something bad not because they've committed a crime or done something offensive to you know the owner of the venue but because they're associated even like really loosely with somebody who's done something to tick off the owner of the venue.
00:59:40
Speaker
I hope that James Dolan and MSG Entertainment get their backsides handed to them in a civil lawsuit. I don't know what the legal foundation is for it, but there's this old Shakespeare quote that comes up in my mind from time to time where he said, I have no idea what it was in. I could probably look it up.
01:00:02
Speaker
I could definitely look it up. Chat GPT knows, I bet. But it says, do as adversaries, do in the law, strive mightily and eat and drink as friends. We got to be able to separate this stuff from the courtroom advocacy, from legal advocacy, and then like, here, let me go destroy your, not destroy your life. Her life's not destroyed. Let me humiliate you in front of your nine-year-old.
01:00:30
Speaker
It's humiliating to be kicked out of this venue for not a good reason. She wasn't drunken disorderly. She got kicked out of this venue because some partner of hers in another state did something that this guy doesn't like. There's got to be some separation there.
01:00:49
Speaker
I mean, I think there has to be, this is all stuff that we're going to see more of because this is only enabled through technology, right? If this was something where they just had like shown a picture, a printout of her from her website, all the attorneys from the firm, right? Obviously it's not just the one that was actually involved in the litigation. So you have to have a printout of all the whole handout and you hand it out to all the security guards. You say, do me a favor at this massive venue with all these people. If you see this woman, make sure she's thrown out.
01:01:15
Speaker
the odds of them finding her and throwing her out are pretty slim. So like this couldn't happen until a few years ago. And so I don't know what the legal basis is for not being permitted to do this, but I think it has to be formed. And I think the only reason it doesn't exist is because it hasn't had to be an issue before now. You know what I mean? Yeah, I mean, this is, you know, private property. Really, this is basically a
01:01:41
Speaker
This is a, you know, a rich jerk story that happens to have a tech angle because if she was, if they caught her through traditional means, it's just like, Oh man, that guy's like, it probably doesn't even generate a headline. Uh, but what I, but it's going to be so easy to be a jerk now. It's going to be so much easier to be a jerk. It's automated stuff.
01:02:03
Speaker
Though I do think that the reaction to this is maybe extremely low tech, anti facial recognition technologies. I don't know. Oh, I'm, you know, masks, for example.
01:02:17
Speaker
um which we were all wearing for a little bit um and some of us still are uh like i mean dang this is where like if you are if you want to go to madison square garden and you uh and you maybe you don't know maybe your firm has sued james dolan right
01:02:36
Speaker
Your brother-in-law is a paralegal at a law firm that once sued him personally for not paying a bill somewhere. You've got to wear a mask now. Yeah. Maybe you're still a cautious individual and maybe just throw one on. Yeah. Or maybe there's going to be higher tech stuff that comes out at some point so you don't have to wear anything or look silly by taping your nose to your
01:03:03
Speaker
I got a lot of ideas. That lawyer should hook me up about all my ideas about circumventing facial recognition technology. It all involves scotch tape in your nose. Yeah, scotch tape. Okay, I'm actually tapped out. Those are my two ideas.
01:03:23
Speaker
Okay, so I think we were probably gonna head into just sort of talking a little bit quickly about what's going on in our lives. Just, you know, recommendations, general things. You get to go first. Sure, I'll go first. So the only thing that's been going on with me lately, really, is I wrote an article that has me being accused of
01:03:43
Speaker
being like a right-wing propagandist. Basically, I wrote an article for... I have a column for Bloomberg Tax, and I wrote an article that has the thesis that we should consider whether or not Trump's tax returns should be released to the public. Basically, what
01:04:04
Speaker
you know, what, what, what, and this serves. And so for those who don't know, I know that's pretty much what I've been. There was a president, um, from 2016 to 2020. He was in home alone too. He's the guy who tells, uh, Kevin McAllister, uh, you know, he, he says down the hall and to the left. And if you notice Kevin McAllister goes down the hall and to the right and he was also, he was lying there.
01:04:30
Speaker
Was he? Yes, he was in Zoolander talking about how big Derek Zoolander is. Or Hansel, I forget, one of those two.
01:04:37
Speaker
Oh, okay. Famed wrestler also, Donald J. Trump, was the 45th president of the United States of America for four years from 2016 to 2020, and he didn't release his tax returns. He kept promising, saying he would, but he said he was under audit. He said he couldn't. There's all sorts of reasons that he gave that were not probably true. Now, through the House Ways and Means Committee, it looks like they're going to be released to the public probably as we record this tomorrow.
01:05:02
Speaker
And so I wrote a column just basically saying I don't really understand. I'm struggling to see the justification for releasing the returns. And I open with saying that I'm not defending President Trump, former President Trump, as an individual or in his presidency. But I don't see what individuals can do, what the public can do with any information found in there. If he committed a crime,
01:05:29
Speaker
the attention and effort should be put towards charging him. I firmly believe that some of the impetus to hold him to account is sort of the pressure is released by doing something like releasing these returns to the public and him being sort of publicly embarrassed.
01:05:47
Speaker
And I don't think that that's a stand-in for moving forward with if the returns contain evidence of a crime, prosecuting him, just like you would prosecute anyone who commits tax fraud. Tax is one of the few areas where it's relatively clear cut. If you took a deduction for something that isn't a valid deduction, that's not a valid deduction. That's it. If you lied on a form, if you did something to falsify records, that's fraud. That's tax fraud. That's it.
01:06:13
Speaker
It's really not something that's subjective, particularly. I mean, there's, you know, there's some play around the edges and stuff, but in general is so difficult, let me tell you.
01:06:22
Speaker
Right, right, exactly. Especially in a technical thing like that. Yeah. And so I just thought it was my overarching sort of thesis. And I had some pushback. I didn't get, I think because of the holidays, the crazy people were sort of busy. And so they didn't have the same sort of time and energy to devote to me. So I didn't get any voicemails threatening my life or anything like that. I got a couple of people on Mastodon accusing me of being like a right wing propagandist or something.
01:06:49
Speaker
You drop your spiciest takes on Friday nights at 5 p.m. before major holidays. Yeah, slip it in under the radar. Nobody knows. Make sure the big guy is happy with what I said and then I move on. Yeah, that's how Bloomberg likes it. Low readership. We know. Yeah, exactly. Yep. They don't want to need no eyeballs. Just get those terrible takes out there and that's it. I just want to be clear. That was irony. That was irony.
01:07:14
Speaker
Yeah. Which one? I think all of it. Yeah. I was going to say, what specifically? Oh yeah, all of it. Please read my column. Please anyone. Please read my column. Thank you. My children are starving. So that's all that's going on with me. How about you guys?
01:07:30
Speaker
Well, I you know, it's Christmas. I don't have anything quite that exciting. I well, it's actually more exciting. Maybe I don't know. I think yeah, my wife bought me a steam deck, which if you don't know, it's like a Nintendo switch, but for PC games and super cool. That's a good if you
01:07:50
Speaker
want to be able to sit on your couch with like your family or whatever and be able to play whatever you want and ignore them. You have something bigger than your phone all of a sudden, something bigger and more capable than your phone.
01:08:06
Speaker
Um, but it's so, it's really well made, uh, and the, the funniest, but the battery life. So it's good. It plays games that you wouldn't think that you could run on a little thing like this on a switch size thing. Uh, but the battery lasts like two and a half hours.
01:08:24
Speaker
which is not great, and it gets so hot. Within five minutes, I turn it on, I can just tilt it towards myself and get a blast of hot air constantly, just right at my face. I actually used it to warm up my hands when it was really cold over Christmas, and by the way, it was like 30 in Florida, so we were just dying here.
01:08:49
Speaker
Uh, but yeah, I highly recommend for anybody that, uh, that plays games on their PC and use it on the spot. But do you know off the top of your head? Like what is it? Is it running Android and is it like a, like an Intel chip? Like it runs Linux. Uh, I think it uses an AMD GPU. Um, yeah, it's, it's very capable for it's, there's a, uh, I think there might be a $300 version and then there's like $405.
01:09:17
Speaker
$500, but it's very capable for the price. They're definitely taking a loss on that one. It'll run Elden Ring. I'm playing a game called Disco Elysium on it right now, which came out like three years ago and is pretty demanding. A Switch couldn't run this for sure.
01:09:35
Speaker
It's funny because every seal of approval. Yeah. Every time I see your name pop up on Steam, it's because you're playing something, some vampire diaries or something like that. Vampire survivors is very different. That's where we see you on Steam. Vampire survivors, I also recommend that's on your phone. That's on iPhone for free. You can play that for free. I don't need to have a big thing for that.
01:10:00
Speaker
This is the really hard hitting stuff here. My recommendation is gonna be, so it's not actually going on in my life right now. It's something that happened in my life recently. Early in December, I took my family to Disney World in Orlando and didn't bother to go see Jake while I was there, but. Yeah, that was really messed up.
01:10:21
Speaker
Yeah, sorry. Yeah, I'm not a good friend. But recommendation-wise, the recommendation is this. The coolest amusement park attraction that I have ever been on is the new Guardians of the Galaxy Cosmic Rewind roller coaster in Epcot at Disney World. It is so cool. I strongly, strongly recommend it.
01:10:44
Speaker
Uh, it's funny how, uh, so my family were big Disney nerds. We shortened it to Disney nerds. I don't think we made that up. Uh, but like Epcot used to be the bummer park that your parents made you go to cause they wanted to drink wine and eat food as they walked around the circle. Uh,
01:11:02
Speaker
And now it's like the best ride at Disney world is there. And now also I'm an old man. I was going to say, Oh, now I'm old. And this is, and this is how I get to drink wine. Yeah. No drinking wine, but definitely stopped and basically tried absolutely every single thing from the bakery in the Norway section of Epcot, which is my people.
01:11:27
Speaker
And so recommendation, go to Epcot at Disney World, ride the Guardians of the Galaxy ride, and then stop at Norway's bakery. I don't know what the name of it is and get yourself a Norwegian school bread. Totally worth it.
01:11:43
Speaker
It's funny, so many of my friends come to Orlando for Disney and they're like, yeah, let's get together. And I'm like, okay, we'll see. And then it's like, you know, they have kids and it's like, oh, planning a Disney vacation is like a wall to wall.
01:12:00
Speaker
drain and just like actually getting out of the Out of the plan is this such a like an imposition? It's like I'm I know Okay, if you can if you find a way to like if you ever find time to break away I know that it's taking up so much your time anyway
01:12:21
Speaker
Yeah, Disney, visiting the Disney parks, not to put you on blast, visiting the Disney parks is like kind of my worst nightmare in a lot of ways. But I'm glad that someone last fall. Yeah. Somebody I'm glad so many people, my wife loves it. My wife loves it.
01:12:39
Speaker
Um, but like the, it's, uh, yeah. Um, I'm glad I get to see so many friends and my wife and my son love it so much, but man, being around that many people is like, uh, the most intimidating thing that I can imagine. Anyway, I'm with you.
01:13:00
Speaker
All right. I think, uh, we are all set and I realized I forgot to, uh, hit the, uh, the intro music. So all this episode we'll have is the outro music, but, uh, I think we could sign off, right guys? All right. Yeah. Thanks for listening to the first episode. This is definitely the first episode. There was never an episode before this. No other episodes exist. We hope you enjoyed the first episode when you deserve that episode audience. When you burned it. Yes.