594: We Just Found It on the Doorstep

32:03And it seems like both John and Federico are getting very wrapped around the axle with regard to AI stuff. And I'm not saying, I don't mean to imply that they're wrong or that's bad, but they are getting ever more perturbed about what's going on with AI crawlers. And I mean, to a degree, I get it. So that was on the 17th of June. John says, here's how you can protect yourself from crawling. And then on the 21st of June, Business Insider writes and says, oh, ha, OpenAI and Anthropic seem to be ignoring robots.txt. And if you're not familiar, if you have a web page or website, I guess I should say, where you control the entire domain, you can put a file called robots.txt at the root of the domain.

32:47So, you know, it would be marco.org slash robots.txt. And any self-respecting and ethically clear crawler will start crawling marco.org, or whatever the case may be, by attempting to load robots.txt and seeing if there's anything there. And if so, there's a mechanism, a schema, if you will, by which the robots.txt will dictate who or really what crawlers should or should not be allowed to crawl that site. And it's by path. They can say everything in this directory, you shouldn't crawl everything here, you can crawl it. So you can sort of subdivide your site to say which parts are accessible. Yeah.

33:27And I have thoughts on that, but we'll come back to that. Yeah. I mean, whenever you're ready to interrupt, to be honest, feel free. Okay. Let's talk about robots.txt. Well, just actually very quickly, I apologize. I gave you the green light, now I'm giving you the yellow light. Just very quickly. I was already in the intersection. it's important to note that robots.txt has never been enforced in any meaningful way it's been kind of a friendly agreement amongst pretty much the entire world wide web but there's never been any real um wood behind the arrow or whatever the turn of phrase what we call advisory yeah like advisory locking it is a scheme that people who agree to that scheme can use that scheme to collaborate and work together.

34:10But there is no actual mechanism stopping anyone from doing anything. It is literally just a text file that you can choose to read or not. Right. So with that said, Marco, carry on. Yeah. And so robots.txt is basically a courtesy. It is a website saying, please maybe follow these rules if you would. But it is not a legal contract. It is not a legal restriction. It is not technically enforced or enforceable, really. It is also not universally used and respected. And I can tell you, I operate crawlers of a sort, and I don't use robots.txt. So when Overcast crawls podcast feeds, I don't even check for robots.txt.

34:55I just crawl the URL as the users have entered them or as they have submitted them to iTunes slash Apple Podcasts. What robots.txt advisories were originally for was not like, hey, search engines, don't crawl my entire site. That's not what they were for. What they were for was mostly to prevent runaway crawls on parts of a site that were potentially infinitely generatable. So things like if you had a web calendar and you can just click that next month, next month, next month button forever if you wanted to. And so a web crawler that indexes a page and then follows every link on that page, if it's hitting a web calendar, it can generate basically infinite links as it goes forward or backwards in time.

35:39So the main purpose of robots.txt was to kind of advise search engines, and it was specifically for search engines. It was to advise them areas of the site that crawlers should not crawl, mostly for technical reasons, occasionally for some kind of privacy or restriction reasons. But usually it was just like technical, like, you know, hey, don't get into an infinite loop, which was largely unnecessary because the web crawlers eventually kind of figured out, like, how to limit things on certain sites. and they eventually made themselves more advanced and that wasn't really necessary anymore, even for that case.

36:14I think the primary use case was keep this out of your search index. Like, I think an anti-decent crawler is not going to get into an empty loop, but keep this out of your search index was, you know, and it was respected by the popular search engines of the day and still is. I think Google still reads robots.x and still respects it. But the thing is, like, the whole idea of, like, well, I don't want any bot to crawl this. that there it was so based on assumptions about search engines in particular web search engines the the current uh drama around trying to apply it to ai training i think it's missing a lot of that context that when this kind of unofficial standard was developed it was all about web search engines and when you think about like how the web search engine dynamic has always worked with web publishers.

37:07There was never really any official contract between anybody that said like, hey, Google, Bing, all the other search engines that have come and gone over the years, crawl my page, go ahead, index it, go ahead, even though technically that is making a copy in your server's memory and might be some kind of copyright violation, doesn't really matter because the purpose of this is going to help me. It's going to make people able to find my page through your search engine and will direct people to my page and I will be able to have them there, make money, maybe have them subscribe to my site in their browser or whatever.

37:40So there was that implied symbiotic trade-off that, okay, I actually, as a site owner, I want search engines to mostly index my site because I want people to be directed to my site from the search engine. And so robots.txt was entirely in that context. It was never anything that was some kind of like legal contract that said, you must obey my rules. That really has never been tested until fairly recently. Like that was never really something that really ever came up. I mean, there have been a couple of things here and there with like Google News and news publishers in certain countries and stuff.

38:15But like for the most part, the basic idea of robots.txt was really just, please, like that's it. It was like, please do this or don't do this. And even then, it was often used in ways that harmed the actual customers using things or did things that were unexpected. This is why I don't use it for Overcast's feed crawlers. Because if you publish an RSS feed and submit it to Apple Podcasts, I'm pretty sure you intend for that to be a public feed. and so I feel like it is not really my place to then put up an alert to my user to say hey this person's robots.txt file actually says disallow star on this one path that this feed is in and so I actually can't do this for you.

39:03That would feel like I would have first of all no incentive to do that and second of all because of its intention and context as a standard for search engine which I'm not this doesn't really apply to me and my use. And there were all sorts of things over the years too, like you could specify certain user agents like, alright, Googlebot, do this. Yahoobot, do this. And that was also problematic over the years too because it disadvantaged certain companies if you just had bad behavior once. Or if a site owner just had one bad thought about one of these companies once and then never revisited it or whatever, then that company was allegedly disallowed from crawling the site.

39:45Why? Well, I mean, it's not even that. It's like, you know, for people that know the technology behind it, you know, don't allow Googlebot. The way you identify Googlebot is by the user agent string, which is part of the HTTP request. And anybody can write anything there. And so all someone had to do was say, I'm Googlebot. And then just write a script that slams a site. And people are like, oh, my God, my site's being slammed by Googlebot. No, it's not. It's being slammed by a thing that put that string into the user agent header. like it's just there's no security no authentication and it's like email anybody and people forget this about email all the time it's that's a real email from santa north pole anybody can write anything i know in email there are technologies to try to make this better but with hdb headers the user agent string there's no security behind that so if you're doing making any decision based on the user agent whether it's a decision to allow something with a particular user agent string or disallow something or you try to make decisions about i'm getting all these hits and i look at the user agent string and the user agent string is x therefore it must be google therefore google is bad you have no idea who yeah who or what that is especially with like proxies and things bouncing around the web or whatever so just like robots.txt it's all just sort of a politeness agreement and convention that only works when all parties involved are being honest and acting in good faith and that is not something that is true broadly speaking on the web right and and when you're looking at like you know legalities or copyright issues as i was saying earlier like none of this has really ever been tested because the way the way it was being used like the you know the deal between search engines and publishers was mutually beneficial and publishers for the most part who were not like you know bad business people like for the most part publishers really wanted to for search engines to index their public content and their private content shouldn't be accessible to the crawlers.

41:40It shouldn't be exposed to the public internet if they want it to be private. Using robots.txt to try to say, I want you to only use the content on my site for this purpose, but not that purpose, but I'm going to keep serving it publicly and making it available to any bot that comes around publicly. You just have to maybe be polite about it. I feel like this is the wrong tool for that job. That job is more of a legal question. Right now, again, we haven't really had much of an agreement between publishers and search engines and other big aggregators before. I think there have been legal cases about it, especially in the early days.

42:19Because in the early days of the search engine, the idea that you were to go to a website that's not yours and type in a search string and see text that came from your website on someone else's website on the Google search results page. I believe there were legal cases about that. And I think the result was that Google is allowed to run a search engine. Right. And some of that can be considered fair use. I mean, it's like, especially the old style search engines where what you'd see is a series of links that are search results and maybe a summary below them before Google started doing the thing where it's like, actually, I'm just going to give you an entirely unattributed snippet at the top of the page that tries to give you the answer you were looking for without sending you to any site.

42:58And of course, that snippet is now powered by their large language models. But before it wasn't, that is still up for grabs. And we'll talk about that in a little bit. But the basic idea of a search engine that indexes the web and allows you to get links to the things that it has indexed, I believe actually has been tested in court. And either way, whether or not it has been tested in court in your country or in the U.S. or whatever, practically speaking, I don't think there are many as much disagreement about the utility of that. People like having traffic sent to them by Google. There's arguments of Google being too big and there should be competition in the search space.

43:31but the concept conceptually a web search engine i think we all agree is a good thing that is necessary and should exist and helps everybody sure but you know if you're going to start making qualifications of like all right well here's how you have to use my content or not use my content robots.txt is not the way to do that that is not any kind of legal binding that is not any kind of technical restriction i i would i would even question whether it's even a good idea to even still have those files these days and to expect to expect anything from them well i mean i think what people are expecting so we'll read this thing from the perplexity ceo in a second but like i think what people are expecting is for the ostensibly good faith actors to do what the existing ones do google honors robot.txt and so do the other things apple honors it with their apple bot thing so do the other things that crawl the web right nobody has to follow it but the good faith actors do and so i think most of the pushback here is hey i did i thought you weren't just a random you know fly-by-night company or a bunch of you know script kitties or whatever i thought you were a big important serious company uh and you you have a crawler that crawls the web and i should be able and you should use a user agent that looks like you right and we you know we won't ban your user agent when someone fakes it and spams our site with it but we'll just say here are the rules for you uh you can't crawl these urls you can't crawl any of our urls you can crawl these or whatever like i think this is a reasonable tool for that job provided you understand that the tool only works if the people on the other end agree and say yes we will honor your roads that text and i think part of the anger is the ai companies are not behaving the way the search engine companies did uh and that's the pushback We are brought to you this episode by 1Password Extended Access Management.

45:27Imagine your company's security like the quad of a college campus. There are nice brick paths between the buildings. Those are the company-owned devices, IT-approved apps, and managed employee identities. And then there are the paths people actually use, the shortcuts through the grass. Those are unmanaged devices, shadow IT apps, and non-employee identities like contractors. Most security tools only work on those happy brick official paths. But a lot of security problems take place on the shortcuts. 1Password Extended Access Management is the first security solution that brings all of these unmanaged devices, apps, and identities under your control.

46:07It ensures that every user credential is strong and protected, every device is known and healthy, and every app is visible. 1Password Extended Access Management solves the problems traditional IAM and MDM can't touch. It's security for the way we really work today. It's available now to companies with Okta, and it's coming later this year to Google Workspace and Microsoft Entra. Check it out at 1password.com slash XAM for extended access management. So once again, 1password.com slash XAM. Thank you so much to 1Password Extended Access Management for sponsoring our show.

46:46so we have a roundup from michael sai that we'll link in the show notes that talks about all this and then uh on the same day on the 21st of june perplexity ceo aravind serinovas responds to plagiarism the plagiarism and infringement accusations so this is a post on fast company and aravind says and this is a quote we don't just rely on our own web crawlers we rely on third-party web crawlers as well. So it's not my fault they did it, right? It's over there. Well, who is over there? I don't know. It's the people over there. So reading from the Post, and this is a direct quote from the Post, but not from Aravind.

47:26Srinivas said the mysterious web crawler that Wired identified was not owned by Perplexity, but by a third-party provider of web crawling and indexing services. Srinivas would not say the name of the third-party providers citing a non-disclosure agreement asked if perplexity immediately called the third party crawler to tell them to stop crawling wired content srinivas was non-committal quote it's complicated he said what is this facebook is it 10 20 years ago anyways srinivas also noted that the robot exclusion protocol in other words robots.txt which was first proposed in 1994 is quote not a legal framework quote he suggested that the emergence of ai requires a new kind of working relationship between content creators or publishers and sites like his so this is actually uh something that a bunch of ai ceos and other big weeks have been doing is basically saying oh well don't ask us we outsource that it's like come on this is like ceo 101 yeah you outsource lots of things right but in the end it's your company whatever you're you know it's like you know if you say oh we outsource it and they're doing something they shouldn't it's like saying you know we outsource to some company that uh makes our bread for us at our sandwich shop uh And the bread's coming back with shards of glass in it.

48:38You don't say, well, don't blame us. We outsource it. Tell your bread maker not to put glass in the bread. Right? Or are you saying you explicitly said, it's okay if you put a little glass in the bread. You have to take responsibility. Right? And so this is not just the perplexity. I've seen like three or four stories where AI CEO says, oh, that's not us. That's a subcontractor. So we don't have any control of that. It's like, wait, what? Like, just own it. Just say, we've decided we're not going to honor robots.txt. Because everyone knows you're not doing it, and you can't try to blame it on a third-party thing or whatever, and then defend that.

49:11And that's kind of where they go into it. It's not a legal framework, blah, blah, blah. And like I said, I think the pushback is not like, you know, they're legally required to do this or whatever. It's just that, like, we thought you were going to behave like a search engine, and you were going to be a polite member of web society. and it's clear that because you're like an ai startup you're like yeehaw cowboy time uh it's a wild west uh you can't fence me in uh we're not acting like google because we don't have to so tough luck and so people are mad at the companies right that there's there's no legal argument here there's no like it's just a decision that they're making and by the way marco on your decision not to do it i would say the closest analog of overcast is that you're a web browser web browsers don't honor robot.txt if you type a url into the address bar of your web browser you expected to load that page you don't expect to load robots.txt and say oh this site says i'm not supposed to load this page that's not what robots.txt is for so if you are a web client used by an individual user like a user loads an rss feed in a podcast that is a single person using a client application to browse the web you know to get an rss feed right that is very different than an automated crawler that is crawling all over the entire web and following links right that's what robots.txt it's for robots unless you are literally a robot while using overcast which i don't think you are it's like overcast should not look at robots.txt because it's not a robot and it's not being used by a robot i have a whole podcast about this well but it's like but if you look at what perplexity is doing it's i think a lot closer to a browser than a search index well so the perplexity's business is complicated because it's a whole question of like uh do people go there to get links out to other places or do they go there to get the answer that you attempt to attribute and i think people will get angry with perplexity when they provide an answer but then don't say where this answer came from and even if they do say where this answer came from they're like you provided too much content this is the same problem people are beginning to have with google is like you're supposed to be sending me traffic you're not supposed to be removing traffic right by either giving an answer that's synthesized from a website and not telling them the source or basically like inlining the entire inlining my entire web page for example and saying you don't need to go to that website here's the whole page right here that is not kind of what people expect out of a search engine so but perplexity is so new and so young and they haven't quite figured this out but just at the crawling stage so people who are seeing their their website crawled and they're going to web you know the perplexity's service and saying oh i can find my content there and I put you in robots.txt and you shouldn't be crawling in this herpesity is like, we don't have to look at robots.txt because that's just an advisory thing and we've chosen to ignore it.

51:54And so people are angry. That's what it boils down to. Well, and I think there's going to continue to be more and more applications over time of technologies like AI summarization and action models and things like that, where some fancy bot basically is going to be browsing and operating a web page on behalf of a user that is kind of like a browser but it's it's a very different form that i think breaks all those assumptions with publishers like this is one thing that i faced when i was making instapaper a thousand years ago you know instapaper would save the text of a web page to read later and only the text not like all the ads and the images and everything like that i was very careful though to not make features that would enable somebody to get the text of a page without having first viewed the page in a browser or a browser-like context.

52:47So it would load the whole page. They would see the page. If there were ads, those ads would load on the page. They would see those ads. And then they could save what they were seeing. And then part of that would be saved in newspaper and shown to them later. And that was always a very tense balance to try to maintain. because what I didn't want was widespread scraping of people's text without loading their ads, but I figured that seemed like an okay tradeoff because that was literally just saving what was already sent to the browser and what the user was already looking at. But a lot of these new technologies...

53:22First of all, I probably wouldn't attempt that today. But a lot of these new technologies, I think, break a lot of those little details. like the like if you if you have some kind of bot that's doing something on a website that's like so you know suppose what it's one of these action models where you're saying all right book me a flight you know all this stupid like book me a trip thing that all of these ai demos from these big companies keep trying to do even though nobody ever wants that suppose you you have a book me a trip kind of thing and with an ai model and the idea is that model will go behind the scenes and will go operate Expedia or Orbitz behind the scenes for you and manipulate things back there to find the best flights and hotels and whatever else.

54:06Well, those sites make some of their money via ads and affiliate things and sponsor placements on those pages. If you have some bot operating the site for you, kind of clicking links for you behind the scenes in some kind of AI context, that bot is not going to see those ads. It's not going to click those affiliate links. It's not going to pick the sponsored listing. It's going to just kind of get the raw data, and that's it. And that will be violating those sites' business models if that happened. That really has not happened at massive scale until fairly recently. So this really has not been challenged.

54:39This really has not been legally tested that much. This really has not been worked out. Like, what are the standards? What are the laws? What are the legal precedents? How much of this is fair use versus not? You know, for the most part, until very recently, we could pretty much just say, all right, if you serve something publicly via public URLs and anybody can just download it, then nothing bad would really happen to you and your business model for the most part if some bot came by sometimes and parsed that page for some other purpose. It wasn't a big deal. But now there's a pretty significant difference in scale and type of replacement.

55:19Now, with a lot of these AI products and with Google Search itself, you know, increasing over time and then more recently rapidly increasing, what we're seeing now is full out replacement of the need for the user to ever look at that page. That's a pretty big difference. And it's really bad for web publishers. And kind of, you know, then consequently really bad for the web in general. We have a pretty serious set of challenges on the web already. even before this new wave of LLMs came by to further destroy the web, we already had a pretty bad situation for web publishers for lots of other reasons over the years.

55:58To have something that removes the need for many people to visit a page at all, that is going to crush publishers. And so it does make sense why everyone's freaking out about this. It makes a lot of sense. I do caution people, though. I don't think it's a very good business move or a very good technology move to say I'm going to just block AI from being able to see any of my stuff because that's a pretty big hammer and that's a pretty big blanket statement. And you can't actually block them anyway. When it comes down to technically speaking you literally can't stop them. Right. Unless you stop everyone from viewing your website in which case you don't have a website.

56:39Right. So I think it is wise to focus on I don't think I did learn anything else. trying to prevent uses of your content that remove the need to visit your page because that is a direct attack on your business model. That makes a lot of sense. I don't think it's wise to say I don't want any AI training or any AI visibility of my page. That I think is probably short-sighted and probably a bit too much of a blanket statement. And that I don't think it's good for any party involved to have that kind of blanket ban on it. I know. Well, people want, though, that what people the publishers in particular want is they want a, you know, a an ecosystem of members who do agree to some rules of politeness and say, look, we should agree on a system that lets me tell you that you shouldn't do X, Y and Z on my site and you should agree to it.

57:35and we'll feel better about you if you do that. And part of the reason I think Instapaper, your example, was not a particularly big problem is like you said, scale. And anything with AI in the name these days, people flip out about it and think, this is going to be as big as Google. Instapaper was not as big as Google, right? No. It did not have billions and billions and billions of users. If it did, if Instapaper had Google scale, I bet there would have been a hell of a lot more scrutiny on even the very conservative things that you did. But because it was small, it's not a big deal. Like that's part of the sort of the ecosystem of the web is there's all sorts of small things that don't have particular big scale.

58:11They're doing all sorts of weird stuff. Nobody cares about them. We allow them to exist. It's fine. But now these big names in AI, AI is the next big thing. You're an AI company. You have a lot of funding. Everyone looks at them and think that could be the next Google. That could be the next thing with billions and billions of users. So we better take whatever weird stuff they're doing way more seriously than we would take Overcast. and even with Google, the current giant in the world of search and they're trying to replace sites and giving answers on the site or whatever. Neil I. Patel coined a term, I think it was his, about this called Google Zero, which is the point at which publisher websites get zero traffic from Google search, right?

58:49Because it's been going down and down over the years because hey, you'd write type of Google search and look, the answer to my question that I typed into Google, it's right on the Google results page. It's unattributed and I don't have to, even if it was attributed, I don't have to click on any link to get to it because the answer is right there and so google has been sending less and less traffic to websites and google zero is when you notice hey you know what you know how much traffic we're getting from google searches zero i don't know if it's absolutely zero for everybody but it's sure going down and it's a scary world to have what was once the massively largest source of your traffic to your website uh disappear um but yeah like whether or not it is wise to exclude to try to to ask to be excluded from pick you know whatever whatever ai crawler thing from whatever open ai perplexity whatever i think most publishers just simply want that choice and to have that choice the the crawlers need to agree because again there is no technical way to stop this short of doing like putting your entire site behind a paywall and even that's not going to stop them because they'll just pay and have their crawler go through it like it's that's the thing about publishing on the web you do it's like drm you want people to see your movie you can't make it impossible to see your movie you have to give the viewer an ability to see your movie movie but once you give the viewer the ability to see your movie they can see your movie like but what if they see it but also record it i want them to see it but not be able to see it can i do that and the answer is no right so if you're publishing on the web you have like it's like anything else that's why margo was right to call this a legal thing like things are published all the time they were published in paper you know like the books or whatever it's like but i can take the book and look at it i can see all the letters in it haha the book is mine well no actually we have laws about about the stuff that's in that law that book we have this thing called copyright and even though you can technically read it and you can technically copy it increasingly more easily over time with technology we have laws surrounding it to control what you can do it and robots.txt people who think of robots.txt as some kind of like technological bank vault it's no more of a bank vault than you could put on a book like you do want people to read it and you can't stop them from being able to copy it and these days it's really really easy to copy a book especially if it's an ebook right setting aside the whole drm thing um what you want is some either in a sort of polite society an agreement among the large parties that actually are significant to get along and then failing that you want laws to provide whatever protections you think are due to you and yeah the google search stuff has i feel like been hashed out probably in the altibus today's but who knows and the ai stuff has not yet been hashed out.

1:01:22And so moving on to this next one, because we have a lot of these items, Microsoft, at least someone in Microsoft, has a very interesting notion of what the deal is on the web and potentially what the law should be surrounding it. So this is a post on The Verge by Sean Hollister, who writes, Microsoft AI boss Mustafa Suleiman incorrectly believes that the moment you publish anything on the open web, it becomes quote-unquote freeware that anyone can freely copy and use. When CNBC's Andrew Ross Sorkin asked him whether AI companies have effectively stolen the world's IP, Mustafa said, I think that with respect to content that's already on the open web, the social contract of that content since the 90s has been that it is fair use.

1:02:03Anyone can copy it, recreate it, reproduce with it. Sorry, recreate with it, reproduce with it. That has been freeware, if you like, and that's been the understanding. Microsoft is currently the target of multiple lawsuits alleging that it and OpenAI are stealing copyrighted online stories to train generative AI models, so it may not surprise you to hear a Microsoft exec defend it as perfectly legal. I just didn't expect them to be so very publicly and obviously wrong. I'm not a lawyer, writes Sean, and that's also true for me. But I can tell you that the moment you create a work, it is automatically protected by copyright in the U.S.

1:02:35You don't even need to apply for it, and you certainly don't void your rights just by publishing it on the web. In fact, it's so difficult to waive your rights that lawyers have to come up with special web licenses to help. This is so gross. Like, I'm not as riled up as a lot of people about people, about, you know, these AI bots crawling my website. Like, sitting here now, I don't find it that off-putting. I don't love it, but whatever. This, though, this is disgusting. So this is such a weird statement, because everybody knows how copyright works. I'm sure this person knows as well. But to say that, like, oh, once you put it on the web, it's freeware, which is a term that mostly applies to software.

1:03:13but like the idea is you can recreate it reproduce it uh you know copy it like no no no like there like that's those are specifically the things we actually do have laws around what we don't have laws around are the more complicated things like oh can i train ai on it or whatever and we'll get to that in a little bit but like it's such a weird thing to say that like oh as everyone knows since the 90s once you put it on the web you forfeit all ownership that's not true at all and i mean that's that's like it's one of the things that's great about the web is oh it's just like books it's printed word right and especially in the beginning it was just a bunch of words and we already have laws surrounding that right and that's why there were cases about search engines like are search engines copying it because you know we got this whole you know giant library of laws about copying text my website has text on it and google's copying it and they had to duke it out and say actually what google's using doing is you know fine within these parameters blah blah right uh but that fight was fought because it was an example of copying but yeah this i don't i mean obviously the the microsoft ai uh leadership this guy is not a lawyer either and and so but like it's that's not how you should defend this you shouldn't defend it by saying uh you know everything on the web is a free-for-all because that's never the way it's been and it's not the way it is now this is a yet another foot-in-the-mouth problem from microsoft i'm not sure what's going on over there but they really need to uh take a lesson from apple and maybe try to speak with one voice instead of having individual lieutenants make really terrible statements to the press yeah Yeah.

1:04:42So Louis Mantia writes with regard to permissions on AI training data from the 22nd of June. Louis writes from John Gruber today on the 22nd of June. It's fair for public data to be excluded on an opt out basis rather than included on an opt in one. And then Louis continues. No, no, it's not. This is a critical thing about ownership and copyright in the world. We own what we make the moment we make it. Publishing text or images on the Web does not make it fair game to train AI on. The public in public web means free to access. It does not mean free to use. Also, whether reposting my content elsewhere is in good faith or not, it is now up to someone other than me to declare whether or not to disallow AI training web crawlers in their robots.txt file.

1:05:25To allow, excuse me, to add insult to injury, that person may not have the knowledge or even the power to do so if they're posting content they don't own on a site that they also don't own, like social media. So this is, he's so close to getting to the crux of this. in the first little paragraph here, he's basically declaring that training AI on your data is exactly the same as copying and reproducing it. And that is not something that the world agrees on. Louis' opinion is that it is. The courts have not yet weighed in. I think to the average person, they would say, are those the same things?

1:05:59Because they seem like they might be a little bit different, kind of in the same way that indexing your content in Google is a little bit different than just literally copying it and reposting it on the website right but anyway if you agree that it's the same as copying then yeah sure but then the second bit is getting to even more of the heart of it here which is like okay so let's say we do agree that it's the same much you know not not proven yet but anyway um what about when somebody like posts a link to your site on a social media network and on that website they do a little embedding inlining of like the first paragraph or whatever like what if someone copies and pastes a paragraph of your thing on another website right even if you had absolute somehow magical technical control to stop ai crawlers crawling your website if people can read your website and quote from it or embed little portions of it or a screenshot or do whatever on other websites of course you don't control those other websites and so if they allow crawling your stuff's going to end up in the google search index in the ai training model or whatever even though you disallowed it from your website and i i would say that for the most part that we also have laws covering can someone take a portion of the thing that you made and quote it elsewhere there's all legal framework deciding whether that is fair use or not and it's complicated and the law is not a deterministic machine as neil patel who i mentioned before is always fond of saying but we do have a legal framework to determine can i copy and paste this paragraph from this thing on this person's site and quote it on my site so i can comment on it yeah in general you can you know can i make a parody of this uh this article uh you know on my website yeah you can there's a whole bunch of things around that that you know that have been fought out in court that we have a system for dealing with but all of those things say the court determines you you sue them and they say actually this person was allowed to quote that snippet right you lost your fair use case because it's pretty open and shut that's fine that just got indexed by an ai training bot because that person's website allows them you know the polite ai bots or never mind again never mind that you can't stop them anyway right that's just the nature of publishing no matter what you do not have absolute control over every single character that you made you do have control over the entire work and the reproduction of the entire work but you don't have control over other examples of fair use and louis saying oh the it shouldn't be like I shouldn't have to opt out.

1:08:25The default should be that nobody can crawl me. I mean, that's just not only is it technically impossible, but like, that's not the way the web has ever worked. It has always been, we're going to crawl you unless you tell us don't. And even the polite ones, you know, they'll read the thing that you said not to do it, but by default, they're going to crawl you. And I think asking for a world where everything you publish on your website is not only not crawlable by the things you don't want it to crawl up but also not able to be quoted by other people is uh clawing back uh rights that we've already decided belong to other people through fair use so then the music industry decided to get involved yeah multi-billion dollar companies are entered the chat as they would say we talked about this before of like hey louis maintain doesn't want people crawling his website what can he do about it he's just one person uh the music industry They have a lot of money.

1:09:19They have a lot of IP. This is where the stuff really starts going down. Yeah, so reading from Ars Technica on the 24th of June. Universal Music Group, Sunny Music, and Warner Records have sued AI music synthesis companies Udio and Suno for allegedly committing mass copyright infringement by using recordings owned by the labels to train music-generating AI models. the lawsuits filed in federal courts in new york and massachusetts claim that the ai company's use of copyrighted material to train their systems could lead to ai generated music that directly competes with and potentially devalues the work of human artists so from the verge article there's a quote from riaa chief legal officer ken dorishow and that quote is these are straightforward cases of copyright infringement involving unlicensed copying of sound recordings on a massive scale Suno and Udio are attempting to hide the full scope of their infringement rather than putting their services on a sound and lawful footing.

1:10:14And again, that was the RIAA chief legal officer. Mikey Shulman, the CEO of Suno, says the company's technology is transformative and designed to generate completely new outputs, not to memorize and regurgitate pre-existing content. Shulman says Suno doesn't allow user prompts based on specific artists. reading from the lawsuit the use here is far from transformative as there is no functional purpose for suno's ai model to ingest the copyrighted recordings other than to spit out new competing music files that suno is copying the copyrighted recordings for commercial purpose and is deriving revenue directly proportional to the number of music files it generates further tilts the fair use factor against it andy baio writes 404 media pulled together a video montage of some of the ai generated examples provided in the two lawsuits that sound similar to famous songs and their recording artists then finally we'll put a link in the show notes to a verge article that discusses what the riaa lawsuits mean for ai and copyright you know i saw somebody say this a few days ago i don't remember who exactly it was but what's going on if the riaa are suddenly the good guys like this is a weird place to be are they though though so here's here's the thing like this is the tricky bit with this and we talked about this with the image generators or whatever so this is significant because they're big rich companies and they're you have to take them seriously when they bring a lawsuit because this is the kind of like who can stop open ai and google and whatever well you know it's clash of titans you need other titans in here to be duking it out right this is i think this needs to this needs to be fought out in in a court in some way i say that before we see what the result will be because maybe the result is not what we want to happen But like, as with the image things, these companies that, you know, you type in a string and they produce a song for you, right?

1:12:04These models are trained on stuff. And these record labels say, yeah, you trained them on all our music, right? Gets back to the question, is training something? Is AI training? How does that relate to copying? Is it just like copying? Is it not like copying at all? Is it somewhere in the middle? Do any of our existing laws apply to it? And we've discussed this on past episodes as well. especially when the company doing the training then has a product that they make money on and as i said with the image training these models that make songs are worthless without data to train them on the model is nothing without the training data this company that wants to make money you pay us x dollars you can make y songs right that's their business model they can make zero songs if they have not trained their model on songs so the question is where do those songs come from if they've licensed them from somebody if they made the songs themselves no problem right again adobe training their like image generation models entirely on content they either own or licensed nobody's mangy about that that's the thing you're doing you own a bunch of images you license them from a stock photo company or whatever you train your models on them you put the feature into photoshop you charge people money for photoshop they click a button it generates an image whether people like that feature whatever legality seems fine these other situations where it's like hey we crawled your site because we don't care about your robots.txt we trained our models on your data on your songs on your whatever right and by the way we have no idea if these companies actually paid for all the songs let's just assume they did they bought all the songs from you know are they uh you know sony music warner records or whatever or they paid for a streaming service they got all the songs they trained their model and then they're charging people to use their model right i've just like the image processing i've always thought that if you're if you have a business that would not be able to exist without content from somebody that you did not pay anything for that is very different than oh we trained an ai model for research purposes or we trained it for you know for some purchase that is that is not like literally making money off of you and this particular case is like okay it's not just that they're making money but the thing they're providing is quote not transformative they keep using that word because that's one of the tests for like fair use is is the work transformative have they taken the thing that existed but made something new out of it and that's they'll argue that in court whether it is or not it's not transformative and also is it a substitute there's another one of the fair use tests is it a substitute for the product is someone not going to buy a drake album because fake drake sounds just as good and they just listen to fake drake right is it a substitute for it doesn't mean does this sound exactly like it that's a whole other sad area of law of like does song a sound too much like song b and they have to pay them whatever when they're all made by humans right this is like you know would someone pay for this instead of paying for this is one a substitute for the other and that's what they'll be duking it out about but i think at its root it is it's sort of like where does the value of this company come from you know like every company has to take inputs from somewhere they you know manufacture something and they sell it to you or they have a service they wrote the software for they pay someone to run the servers and they sell it like there's a there's sort of a value chain there and a lot of these companies are like we would make more money if we don't have to pay for the things that make our product valuable so we don't want to have to license all the music in the world but we do want to train an ai model on all the music on the world so that we can make songs that sound as good as all the music in the world but we don't want to have to pay for any of that and that seems to be not a good idea for from my perspective and this is one of like the different ways you can look at this moral ethical legal i think one of the what the frameworks that i've falling back on a lot is practical if you know for any given thing say if we allowed this to happen would it produce a viable sustainable ecosystem like would it produce a market for products and services would it would it be a rising tide that lifts all boats or would it like burn the forest to the ground and leave one tree left in the middle right you know what i mean like that practical approach people like to jump on like we talked about before with uh vatici and max stories and everything like they want to go to the moral and ethical thing they're stealing from us it's our stuff they have no right and even what i was saying before like oh they're they don't want to pay for this stuff but they want to make money off of it or whatever but practically and this is not the way the law works but this is the way i think about practically speaking i'm always asking myself if this is allowed to fly what does this look like fast forward this what you know is this viable right what if if everyone's listening to fake drake does drake the next drake are not able to make any money does human beings making music become an unviable business and all this is just an increasingly gray soup of ai generated stuff that loops in on itself over and over again right like where are the you know and we have the same thing with publishing on the web like does google destroy the entire web because no one needs to go to websites anymore they just go to google right unfortunately when these cases go you know to court no one is thinking that that's not how the law works again law is going to be is this fair use or whatever does does congress pass new laws related to this or whatever but what i really hope is that the outcome of all these things and the thing i'm always rooting for is can we get to a point where we have an ecosystem that is sustainable which means it's probably you whatever they're suing for is like i think they want like a hundred and hundred fifty thousand dollars for every song or something that is not a sustainable solution you can't train an ai model when you pay$150 ,000 for each song that you trained it on because you need basically all the songs in the world.

1:17:51That's a big number. That's stupid. We do want AIs that can make little songs, right? I think that is a useful thing to have, right? So we need to find a way where we can have that but also still have music artists who can make money making actual music, setting aside the fact that the labels take all the money and the artists get barely anything anyway, which is its own problem. Yeah, separate issue. Right, and there was a good article about that recently about how the labels, Spotify, and the artists and the terrible relationship there that screws over artists. Anyway, I think, I really hope that the outcome of this is some kind of situation where there's something sustainable.

1:18:30There's like, I keep using ecosystem, but it's like, you know, you have to have enough water, the whole water cycle, this animal eats that animal, it dies, it fertilizes the plant, like the whole, you know, a sustainable ecosystem where everything works and it goes all around in a circle and everything is healthy and there's growth, but not too much and not too cancerous. And it's not like everything is replaced by a monoculture and only one company is left standing and all that good stuff, right? But right now, the technology is advancing in a way that if it's not, if we don't do something about it, the individual parties involved are not motivated to make a sustainable ecosystem, let's say.

1:19:06I mean, that's kind of what the DMA is about in the EU. and these AI companies definitely are not motivated to try to make sure they have a sustainable ecosystem. They just want to make money. And if they can do it by taking the world's music and selling the ability for you to make songs that sound like it without paying anything to the music that they ingested, they're going to try to do that. Yeah, it's... I don't know. It's all just so weird and gross. And it's hard because I don't want to be old man who shakes fist at clouds, right? and it seems like AI for all the good and bad associated with it is a thing it's certainly a flash in the pan for right now but I get the feeling that where blockchain and bitcoin and all that sort of stuff was very trendy but anyone with a couple of brain cells to rub together would say, eh, that's all going to fade or it's certainly not going to work the way it is today I think there's a little of that here, but I get the feeling that this is going to stick around for a lot longer.

1:20:09And I think that there needs to be some wrangling done, some legal wrangling. And I get the move fast and break things mentality of these startups that are doing all this. But I don't know. It just feels kind of wrong. Like, again, I'm not nearly as bothered by it as some of our peers are, but it just doesn't feel right. and it definitely doesn't feel sustainable like practically that's true like if like regardless of how you feel about right or wrong if you say if we just let them do this like and these you know these models get better and better and produce more and more acceptable content you can see that it's taking again regardless of how this lawsuit ends up with the whole record labels you can see that it is taking value away from human beings making music and pushing that value to models making music but those models are absolutely worthless without that human generated music at least initially right again maybe in the future there will be models trained entirely on model generated music but then you have to trace it back to where that model get trained like in the end these models are trained on human created stuff and there's there may not be enough officially licensed human created stuff to train them on at this point i think that's you know i think we we want these tools they are useful for doing things even if you think oh they make terrible music sometimes people need terrible music right sometimes people just need a little jingle they can describe it they want it to be spit out right by most people's definitions all of my music is terrible music they do useful things like unlike you know cryptocurrency which does a very very small number of useful things that is not a general purpose the you know ai models do tons of useful things apple's building a bunch into their operating systems you know people use them all the time they do tons of useful things right we should find a way for them to do those things without destroying the ecosystem i think we can find a way for that to happen if you look at the awful situation with like spotify and record labels and music artists that's a pretty bad version of this and yet still it is better than spotify saying we're going to stream all these songs for free and not pay anybody right i wish i could find that article for the notes i'll try to look it up but but even that even that is better than than the current situation with ai which is like we're just gonna take it all for free come sue us and they say okay we are suing you and they'll battle it out in court but like either way this decision goes with the music thing it could go bad in both directions because if they say oh you're totally copying this music all ai training is illegal that's terrible that's bad we don't want that right and if they say no it's fine it's transformative you can take anything you want for free that's also bad so both extremes of the potential decision that a court could make based on this lawsuit are really bad for all of us for the future so that's why i hope we find some kind of middle ground like again with spotify they came up with a licensing scheme where they can say we want to stream your entire catalog of music can we figure out a way to exchange money where you will allow that to happen legally uh and they came up with something it's not a great system they came up with again And if I can find that article, you can read it and see how bad it is.

1:23:15But they didn't just take it all for free, right? And they also didn't, the music labels didn't say, okay, but every time someone streams one of these songs, it's 150 grand, right? That's also not sustainable. So obviously they're staking out positions in these lawsuits and they're trying to put these companies out of business with big fees or whatever. But yeah, this is like, it's scary. It's scary when Titans clash. And I do worry about how the results of these cases are going to be. But I think, I either think we have to have these cases or, and I know this is ridiculous in our country, or we have to make new laws to address this specific case, which is different enough from all the things that have come before it that we should have new laws to address it.

1:23:52And it would be better if those laws weren't created by court decisions. But our ability and track record for creating technology related laws for new technology is not great in this country. So there's that. Yeah. And then it continues because Figma, a popular, I don't know how to describe this like a user interface uh generation tool design yeah design tool thank you uh they pulled their ai tool after criticism that it blatantly ripped off apple uh apple's weather app so this is the verge by jay peters figma's new tool make designs lets users quickly mock up apps using generative ai now it's been pulled after the tool drafted designs that looked strikingly similar to apple's ios weather app in a tuesday interview with figma cto chris rasmussen I asked him point blank if Make Designs was trained on Apple's app designs.

1:24:42His response? He couldn't say for sure. Figma was not responsible for the training AI models he used at all. Who knows who trained it? It's just our model. We don't, I don't know. Do you know who trained it? I don't know. Does anyone know who trained it? We just found it on our doorstep and this is a model. With the real trainer, please stand up. Quote, we did no training as part of the generative AI features, Rasmussen said. the features are, quote, powered by off-the-shelf models and a bespoke design system that we commissioned, which appears to be the underlying issue. So if you commissioned it, then you should know!

1:25:14We had someone else do it, and they gave it to us, and we just took it, and we were like, we didn't ask too many questions. We're just like, it's fine. Whatever you got, just give it. It's probably fine. The key AI models that PowerMake designs are OpenAI's GPT-40 and Amazon's Titan Image Generator G1, according to Rasmussen. If it's true that Figma didn't train its AI tools, but they're spitting out Apple app lookalikes anyway, that could suggest that OpenAI or amazon's models were trained on apple's designs open ai and amazon didn't immediately reply to a request for comment this is seriously the like the spider-man pointing at other spider-man's you know image it's just it's not my fault it's their fault well it's not my fault it's their fault oh no no no no it's not my fault it's their fault was that company i think it was open ai or whatever the sora model that does makes makes movies essentially that someone who's uh responsible for that i was asked in an interview did was your model trained on youtube they're like they didn't give an answer like maybe i don't know listen if you run an ai company figure out how and where your models were trained i'm not i don't know what you like maybe you train them on good things bad things whatever but have an answer don't say we don't know we someone else did it we like like this seems like table stakes like you should know where and on what your model was trained i'm not They're not like granular, like every single individual thing, although ideally that would be great, but it's too much.

1:26:33I get it, right? But when someone says, hey, did you train on YouTube? You should be able to answer that with a yes or a no, right? Not weasel about it. And this one, was this trained on Apple's apps? I mean, anyone looking at it is going to be like, well, if it wasn't, this is the world's biggest coincidence because it looks just like Apple's app. As Gruber pointed out, right down to the really weird like line chart that I never really understood until I saw it explained in Apple's weather app, right? It was obviously trained on Apple stuff. But you have to have an answer, right? If you don't have an answer, say, I don't know, but I'll find out for you and then come back.

1:27:03But like the bar is real low here. Anyway, same situation, different thing. Images, songs, text, UIs, a mock-up tool that makes UIs. It's based on a model. That model is worthless without being trained on a bunch of UIs. Where are you going to get enough UIs to train it from the world of UIs that we take essentially without permission? Is that okay? if we sell that as part of our application is that okay i mean i i wrote a big post about this what in january like excuse me i made this yeah and i we talked about it on the podcast before and i took a while to write this because actually speaking of neil patel i was listening to the decoder podcast and there was an episode where i was debating with somebody about the new york times lawsuit at the time like new york times was suing some company that trained its ai on the new york times and they said you can't do that and going back and forth about like well the model is just doing what a person would do and it's learning and blah blah blah and is the person the same as the model does the model have the same rights as a person and i was trying to write up something related to that and as usual writing helps me clarify my thinking but it is a fairly complicated circuitous route to sort of really dig down into that thought to get to what's at the the heart of it uh and i wrote this thing and i think i did get to the heart of it as far as i was concerned but it's complicated so every time i try to like summarize it on the podcast i find myself like tongue-tied and rather you know you just quote from the paragraphs like i think if you read the post my thoughts are in there but a lot of people have read it and like no one has commented on it so maybe i'm doing a poor job communicating it but uh i was coming at it from the other angle we talked all about training data in this section of the show here i was coming at it from the angle of like um what was then one of the hot topics which is say i use one of these tools say i use the figma tool to generate a ui i use the song tool to generate a song or whatever um that thing that i made what is the legal ethical moral practical ownership deal with that if i use figma to make that auto create ui thing and it makes me a ui and i put that in my app do I own that UI?

1:29:19If I make a song with the song making tools, do I have the copyright on that song? There's been legal cases about this. And I think the only ruling we have now is something like if you make it with an AI generator tool, you don't have the copyright on it or whatever. But the reason I got to that because I was getting with the whole like, oh, you know, training is just like what a human would do. They read all these articles in the New York Times, then you ask the human the answer and they read all those articles and they know, have the knowledge from reading those articles and they give you an answer.

1:29:43Well, that's just what our AIs are doing. I'm like, yeah, but a human is a human and AI is an AI. And is that really what the root of the thing it is? And I kept chasing that down, chasing that thought down and got to sort of the, uh, the thing that confers ownership, right? Like when you make something, it's yours. You, you write something on your blog, you have the copyright to it because you created it. It's so clear, right? What if you draw a picture on a piece of paper? Okay. You got the copyright on the picture, right? What if you use Photoshop to make a picture? Well, now you use this software tool written by a bunch of other people just plain old photoshop not like ai generator like photoshop 3.0 right with layers now uh you use photoshop but you didn't write photoshop a bunch of people wrote software to make photoshop then you then paid adobe for then they gave you that software product you use photoshop to make a picture but still we say well you made that picture you have the copyright on it you are the creator you own it right then we say all right but what if you can't draw what if you what if you tell something like i can't draw i here's what i want i want this picture of whatever example i gave a thing a polar bear riding a skateboard but i can't draw so i asked somebody else say can you draw me a picture of a polar bear riding a skateboard so someone goes and they draw a picture of a polar bear riding a skateboard at that point the person who drew it owns it maybe they use photoshop maybe they don't they own it because they created it they drew it right but then you say okay this was a work for hire i'll give you 10 bucks and our contract says i give you 10 bucks you give me the polar bear drawing now i own the polar bear drawing because i paid you for it that is a market for creative works someone was an artist they i can't draw they could they drew it they asked for money i gave them money they gave me the ownership of the polar bear drawing the copyright is now mine right and the act of creation is clear the person who drew it they created it i paid money for it they sold me their creation now i own it all you know normal right now i say make me a picture of a polar bear on a skateboard but i don't say it to an artist i say it to an image generator it's the exact same thing as i did before before when i did it it was clear that i don't own anything until i pay for that right now when i do that exact thing but instead of typing it into an email to an artist i type it into an image generator and i get an image back who created that image i didn't create it but if you're going to say i didn't create the one that the artist drew for me because you just told the artist what to draw but you didn't create it well if i didn't create that one i certainly didn't create this one because i literally did the same thing i just typed the text in a different text field could literally be the same text it'd be an ai prompt emailed to an artist or sent to an ai so i'm not going to say that i am the creator of that the ai model can't be the creator because computer programs can't own things they don't have they're not they're they don't have rights computer programs are made by people who have rights just like people made people who wrote photoshop they have the rights to photoshop and so on and so forth but the people who wrote photoshop have no rights to the things that people made with photoshop despite adobe's little snafu with their license agreements recently which they clarified but anyway so i didn't make that picture of the polar bear the large language model didn't make it who who owns that picture of the polar bear based on the act of creation where is the act of creation there how did that model create the polar bear well it created the polar bear picture because it had been trained on tons of other images that maybe were or weren't licensed but still i'm looking around of like if ownership is conferred by the act of creation and there's no active creation here what the hell are we what what's what's going on here who owns the picture of the polar bear and that like every time i dig down into some kind of like oh ai is allowed to do this and you're allowed to train it's just what people do or whatever and computers aren't people i always go through to looking for how we confer ownership of stuff like this how we confer ownership of intellectual property how we exchange money for intellectual property how the market for intellectual property works and none of the existence systems make any sense in a world where i can say the same thing to a human and a generator uh that is clearly not me creating anything and yet i do get a picture out of it that came from somewhere and there's no like there's no human actor it's an indirection right and so i think we need new ways to think about a new laws for that type of indirection to say what is the chain of ownership here it's kind of like not quite the same thing but remember the whole thing where like the the like uh monkey took a picture of itself with the camera do you remember oh yeah it was like a camera set up in the jungler whatever and a monkey comes up to it and snaps a picture of himself and the photographer is like well it's my camera so i own the copyright to the picture like well doesn't the monkey own the copyright because it took the picture right and it's like but but the monkey can't own the copyright it's not a person right and believe me a monkey is way closer to a sentient being than an lom right it's like it's a real living thing no one's gonna argue in court that a monkey is not alive and Well, does it have legal rights?

1:34:32Well, I would say a monkey has more legal rights than a large language model, which is just a bunch of numbers in memory. Right? And so this is the kind of conversation we're having. And honestly, this would be so much easier to have if we had actual artificial intelligence as in sentient artificial beings. But we don't. That's just science fiction. Large language models are not anywhere close to that. That would be so much easier because you'd be like, well, conscious beings have rights and we need the, you know, whatever. they always have names of this in sci-fi movies the the ai consciousness act of uh 2732 that gives rights to the ais uh to avert a global war and plunge us into the matrix apocalypse you know what i mean like it's so much easier when you say well people have rights and computer programs that are basically people have rights and it's straightforward but we're nowhere near there so now we're arguing about monkeys if they have the copyright pictures and we're arguing about huge matrices of numbers whether they can create anything or you're saying like oh you're saying basically like the people who wrote photoshop own every picture that's made from because like well no the lm doesn't own it and the person who wrote the prompt doesn't own it but you know who does own it open ai because they wrote the program that crawled all the pictures in the world that trained the model that you paid to use none of those answers are satisfactory in any way like it doesn't feel right it doesn't seem right it doesn't seem sustainable and yet we do need some kind of answer here even if the answer here is that anything again like that one law precedent we had is like if you make something out of ai you don't own the copyright on it it is not copyrightable nobody owns it it's garbage it's slop it's a thing that exists but nobody can claim that they own it so it is free for anybody to take and do whatever they want with but you certainly can't like sell it to someone because you didn't own it it's very confusing i know that I haven't made this any clear.

1:36:17You can try reading my post to see if it becomes any more clear. But really, this is a dizzying topic if you think about it for any amount of time. And I think a lot of people are doing a lot of feeling about it, which makes perfect sense. And honestly, it is more straightforward to feel things about it than it is to think about it. Because thinking about it gets you into some weird corners real fast. It's just, it's a mess. It's a mess. And I don't know what the right answer is, right? Like, it's so gray from top to bottom, and I just, I don't know. I just don't know. Well, and I think we're going to have to be fighting this and working this out for a while.

1:36:57I mean, look at how much disruption to existing businesses, existing copyright law, and existing norms was caused by the web and then the rise of other things on the internet. Like, this is just, this is how technology goes. There are massive disruptions to what has been established, what many people have held dearly. There's massive disruptions to that when new tech comes around sometimes. And sometimes it takes a decade or two to really settle out and work out what are the norms? What should the laws be? What does copyright mean in this new world? Things like that. That takes a long time to work out sometimes.

1:37:36The rise of these AI techniques and models is potentially as disruptive to existing business models and norms and perspectives as the web was when it first came out a thousand years ago. So I really think we're in for a while of just not knowing there's going to be a lot of damage and destruction along the path to get from where we are now to kind of where things settle out. It will destroy a lot of businesses and it will make it hard for a lot of people to do what they've been doing. It will also create a bunch of new businesses and create a bunch of new value and new opportunities just like any other massive disruption.

1:38:18I think this is a very large disruption, and it's mostly only going to start to become visible of what the other side looks like just after a bunch of time has passed and we've gone through a bunch of messiness. And we're in such early days, it's really hard to know where we're going to end up right now. I feel like this is going to be, in some respects, not all, but in some respects, even more disruptive than the initial web. Because the initial web was kind of like text. We have laws governing that. it was a massive shift of wealth obviously newspapers go out of business craigslist gets rich you know what i mean like like but we saw that giant shift paper magazines like the shift of publishing right and web search and doing all that or whatever but during that entire thing people were upset and it was a big turmoil because it was like these things used to be huge every city had 25 newspapers a newspaper reporter was a big job and you know it's like and all of a sudden all that money's going elsewhere to these dot-com things or whatever but during that whole process there was mostly agreement that like newspapers own what they publish websites on what they publish like we have existing copyright laws for this there's the whole google search index thing that we can figure out and you know fair use on the internet and stuff but in general it was just a massive shift of power and money from older industries to newer ones mostly following along the shape of laws and ideas and morals and ethics and societal understanding about the written word, mostly in the early days of the web, especially before social media really came and mixed that up a little bit, right?

1:39:46With the whole aggregation of humans all talking to each other and quoting things and linking out or whatever, right? That, in hindsight, that seems much less disruptive than this AI stuff, which is like, it's a free-for-all. No one knows anything. No one knows what's legal, what's not, what's sustainable, what's not, what should we do, what can we do, what are people doing, how valuable is this, how useful is it, like, just so many questions and we like all the laws that we have that seems like they could apply to this and some of them do apply it's like yeah but there's these huge areas where it's like here be dragons on the map they draw the big dragon and the thing is like nobody knows what's there and there's a lot of money behind it and a lot of people running in that direction and it's not even clear where or how this will shift the power like in the internet in the early days it was pretty clear paper newspapers power is going away from them and towards websites like that trend was visible to anybody with a clue and it was just a question of how fast how hard you know whatever here is this going to shift power massively to the record labels because they own all the music for example or is it going to destroy them because everything they have is now worthless because ai models can be trained on it and it's a perfect substitute for what they previously made and no one wants anything like you can't even tell which direction it's going to go at this point it's so early and i just don't think that was true of the web so this is an exciting time to be alive in many ways, especially if you're in any industry, any creative industry that involves intellectual property that AI touches at all.

1:41:09And at this point, that's nearly all of them, right? And right now, what it does is not, you know, not particularly amazing, but it is good enough for so many use cases. And this stuff generally doesn't get worse over time. Thank you to our sponsors this week, 1Password and Photon Camera and thank you to our members who support us directly you can join us atv.fm slash join members get a bunch of perks including ATP overtime this is our weekly bonus topic that's an extra segment that only members get to hear ATP overtime this week is going to be about a rumor reported by the information and Mark Gurman about some changes and plans to what Apple is going to be working on for the next Vision Pro and kind of what they can maybe do to make the next Vision Pro cheaper and how they're going to possibly do this and everything.

1:42:05That's what we're talking about in ATP Overtime this week. Join now to listen, ATP.fm slash join. Thanks everybody and we'll talk to you next week.

1:42:17Now the show is over. They didn't even mean to begin because it was accidental. Oh, it was accidental. John didn't do any research Marco and Casey wouldn't let him Cause it was accidental It was accidental And you can find the show notes at atp.fm And if you're into Mastodon You can follow them At C-A-S-E-Y-L-I-S-S So that's Casey Liss, M-A-R-C-O-A-R-M-N-G, Marco Arman, S-I-R-A-C-U-S-A-C-R-A-C-U-S-A. It's accidental. Accidental. They didn't mean to. Accidental. Accidental. Tech podcast so long. not so real-time follow-up on my earlier statement about uh apple silicon max not being able to use pci breakout boxes that is not true you can use thunderbolt pci breakout boxes obviously you can't yeah it's just not gpus yeah but you can't use gpus internally either that's the thing yeah so still apple should have put the mac pro in the configurator oh and i suppose they could have said hey do you use pci no one's buying it except you you can use pci cards well you can buy a mac studio and also this third-party product that we don't even sell or you could buy a mac pro which is a product in their lineup i i think two things are simultaneously true number one they should keep making the mac pro because it does have uses and number two absolutely nobody should buy the mac pro effectively like anybody who's going to a page on apple.com saying what mac should i buy none of those people should buy it none no they the app they should that should be like the whole point of this is which it's a path that leads to all of our products and maybe there's only one very lonely overgrown path that leads to the mac pro but it's got to be there look i would say number three your product chooser should let you choose from any of the products depending on which things you answer put as many scary questions in there as you want there's just got to be a path that lands in the mac pro because otherwise like look what they're saying with this is no one should buy this product and i don't think apple believes that if you ask them they said well some people should like okay great but you have a tool that lets people choose and it has every single mac you sell except for that one that just seems like a bug to me someone should report it they should fix it you should report it i just wanted to make a mac pro that's worth buying i mean that's a bug maybe they're working on that we'll say i don't know so what i mean i i feel like we covered this in the past but what what are you waiting for like what what would make it worth i mean am i waiting for anything in particular i like i don't know like because again with the gaming situation on apple silicon max is entirely unclear if i did buy a mac with a big beefy GPU like bigger than a Mac Studio GPU that would be a speculative purchase it would not be like my current Mac Pro which I literally knew I could run Windows games on and do and they work fine and I run literally put into Windows like that's not speculative that's a thing right if I buy if I decide hey I want a bigger than Mac Studio GPU in an ARM Mac I'm like crossing my fingers that some magical point in the future I will be able to do interesting gaming things on it I don't know if I'm going to make that speculative purchase I don't know if Apple's going to make a Mac with a better than Mac Studio GPU in it.

1:45:40And maybe they make it and it's just too rich for my blood and I can't spend that much money on something speculative. Right. Like I said, my default is an M4 Mumble Mac Studio is potentially the computer I replace this with whenever they release that like next year or towards the end of this year or whatever. But I would like to see, you know, show me something. Show me the Mac Pro. Show me something that's not a Mac Studio and giant cavernous case. Right. That's what I would like to see from them. and then i can decide is it worth it for me to get that because it's not a slam dunk like the well it's not as big a slam dunk as 2019 was because again that's just not speculative but it's just kind of wishful thinking at this point to think you're going to be running windows arm games natively on you know you're going to be booting windows for arm on your you know apple silicon mac pro or you're going to be running uh windows uh caliber games in mac os because apple will have gotten all the triple a game developers on board that is all just twinkle in someone's eye right now.

1:46:34It is not a real thing. I just, I feel like, and I'm going to say this, and I know and I understand why it's not appealing to you, but I feel like so many of your problems, well, maybe not even problems, but so much of your life would be so much better if you would just get a damn Mac Studio and a damn Windows PC. And I get it. I don't want to run Windows anything. I don't. And I know you are even worse than me in this capacity. but like that would make so many things so much better in your life i would probably have a gaming pc if i had a place in the house for it but i don't so i mean i hate to break it to you but i really don't think that there is ever going to be a mac pro that does the things that your current mac pro does and i mean that that may be true like i'm rooting for it but like right now the outlook doesn't look so great yeah i would definitely not hold your breath on that i mean like the thing is it's actually kind of if i thought like two years ago like predicted how this would go actually i'm kind of surprised at how much motion there is here the copilot plus pc how hard microsoft is pushing into arm pcs after doing such a bad job with windows rt right apple with its whole game porting toolkit like both those parties both microsoft and apple are actually surprising me with how hard they're trying to make my dream happen they're just not succeeding right but they are they're trying more than i thought they would right i did not i didn't think they'd be like both on both sides i have been pleasantly surprised by the additional effort that they are putting in i think everyone kind of is it's just like i just they're just not really doing it right but but i give i give them kudos for the effort um if i had to pick one thing like i would i would wish that microsoft would commit to a transition to arm but that's not what they want to do they they seem to think that they're going to have a they're going to support x86 and arm forever off into the future which i think is a dumb strategy but that seems to be what they're doing and that doesn't help me and that doesn't help windows games get ported to arm that all that does is bifurcate their market and say well all the all the triple a games will still be on x86 with nvidia cards and arm will just free for people's laptops and microsoft may be perfectly happy with that but it doesn't help me over here with apple silicon i mean what what pc games are you playing with regularity right now that's i don't know if you know this but destiny runs on pc well right but that's the thing like is there no other appliance that you can buy to run destiny can't you do it on playstation destiny runs at higher resolution and higher frame rates on gaming pcs i don't really play it on my mac pro i play it on my playstation 5 for a variety of reasons but it does run better as defined by resolution and frame rate even on my mac pro than on my playstation maxes out at 60 frames per second right and i can get higher than that um depending on settings and you can go way higher you can go i've played it i actually have played destiny on my ps5 at 120 frames per second on my tv but it has to lower the quality substantially and i generally don't play destiny on my tv because it'll burn it in right but i did try that just to see what it was like at 120 frames per second is good all the like the destiny streamers who are out there playing destiny they occasionally have their frame rate displayed in the corner they're triple digits always hundreds of frames per second sometimes pushing them into 200 uh it makes a difference it it looks and feels smoother especially in pvp like that's you know and even if i'm playing on a controller because at this point sadly i'm better with a controller in destiny than i am with mouse and keyboard uh and also controllers way better for my rsi so i'd be doing it anyway uh but yeah destiny is one choice and there's games come out all the time and they come out for pc they don't come out for the mac until three years later when apple puts than a keynote right so you know there's past games there's future games there's my gigantic steam library that i still haven't played through you know i mean i'll get a playstation 5 pro i'll get a playstation 6 i do like consoles they're great maybe someday the gap between pc and console will be diminished even now i would say it's more diminished because 60 frames per second on ps5 is such a change from 30 on the ps4 that i feel like the gap has narrowed because destiny players were playing at 100 you know 200 frames per second back when i was playing 30 now they're playing at 100 200 frames per second i'm playing at 60 right i'm gaining on them so maybe at some point i'll be like you know what i don't need a big gpu and i'll just get a mac studio and be happy with it and that's looking like the most likely situation uh right now but uh you know we'll see i mean typically you're as much as i'm giving you a hard time i i want you to to have what you want i i can make an argument even i can make an argument for the mac pro for a really beefy mac pro that's useful for people that work outside of a music studio like i'm not saying that your desires or wants as much as i'm giving you grief about it i'm not saying your desires or wants are unreasonable i don't think apple will uh will be uh achieving them but i don't think they're unreasonable i mean it's exciting that they did with 2019 because again i've said before like i don't you know despite my gaming things this is not a purchase there's not a rational purchase in the same way that you don't need a Ferrari to get to work faster.

1:51:33People just like fast cars because they like fast cars. Right, and that's allowed. I just like powerful computers because I like powerful computers. It's exactly the same thing. Trying to justify a Mac, me trying to justify a Mac Pro is like someone trying to justify a Ferrari. It's like, well, I need a car this fast to get to my work. No, you don't. Nobody does. But people want them because they're cool. Right. And that's fair, and that's totally fair. But I feel like, to my eyes, we're starting to cross from, from, oh, it's kind of adorable that John is still rocking his Mac Pro, to like, man, I kind of want you to move on to a Mac Studio, because I think you might enjoy it a lot more, you know?

1:52:09Well, I mean, I'm not buying an M2 one at this point. Well, that's fair. No, that's fair. This is not the time to buy a Mac Studio. It isn't. It isn't. Hanging in there for the M4 one. Yeah, I think when the next one comes out, I think that's your move. I can't, I just cannot see a future in which they make the Mac Pro that you want. And so you might as well get the Mac Studio, which is the Mac Pro without slots like that is the new mac pro i can't say it enough and and with a with a wimpy or gpu but they they just like the mac studio is the mac pro they should have called it mac pro that is the apple silicon mac pro they should not have no can you imagine the aneurysm he would have had it doesn't make sense they sell a thing called the mac pro that's way bigger but it's the same computer it's just a built-in pci breakout box i know i know it's still it's still got the slots is still anyway well we'll see how good and by the way by the time i do get this like my computer is essentially five years old now already this is a pretty good run for a computer that i bought at the you know just before the the you know processor transition right which is you know unfortunate for the we already said when it happened like oh my poor mac pro whatever but i love this machine and i've already gotten five years out of which granted is half of what i got out of my last mac pro but you know processor transition right so if i ditch this machine at six years old that's longer than any of your laptops lasted right it's still a pretty good run hey we're just excited if marco makes it six months much less six years he's been pretty good with 16 inch i think it's almost two years old now right no it's the m3 max it's uh it's the black one that came out of mine sorry i mean to be honest lately i haven't been much better so i shouldn't be casting stones in this glass house but generally speaking marco is much more frequent on on his purchases so i mean like no matter what like i feel like i've gotten a good run out of this mac pro and i'm enjoying it for you know as long as like i'm excited that sequoia runs on it that's cool next year probably not right so it's really putting a deadline on this like i said i'm willing to run this with last year's version of the operating system for some period of time if i have to wait right but you know well we'll see what happens like i'm you know i keep my cars for a long time i keep my macs for a long time Beep beep beep.