But what Perplexity is doing when they crawl my content in response to a user question is that they are decreasing the probability that this user would come to by content (via Google, for example). This is unacceptable. A tool that runs on-device (like Reader mode) is different because Perplexity is an aggregator service that will continue to solidify its position as a demand aggregator and I will never be able to get people directly on my content.
There are many benefits to having people visit your content on a property that you own. e.g., say you are a SaaS company and you have a bunch of Help docs. You can analyze traffic in this section of your website to get insights to improve your business: what are the top search queries from my users, this might indicate to me where they are struggling or what new features I could build. In a world where users ask Perplexity these Help questions about my SaaS, Perplexity may answer them and I would lose all the insights because I never get any traffic.
Google has been providing summaries of stuff and hijacking traffic for ages.
I kid you not, in the tourism sector this has been a HUGE issue, we have seen 50%+ decrease in views when they started doing it.
We paid gazzilions to write quality content for tourists about the most different places just so Google could put it on their homepage.
It's just depressing. I'm more and more convinced that the age of regulations and competition is gone, US does want to have unkillable monopolies in the tech sector and we are all peons.
[1] https://developers.google.com/search/docs/appearance/snippet...
Customers liked the in depth value of that content and it converted to buys (we sold experiences and other stuff, sort of like getyourguide).
One day all of our content ended up on Google "what time is best to visit the Sagrada Familia" and you would have a copy pasted answer by Google.
This killed a lot of traffic.
Anyway, I just wanted to point out that the previous user was a bit naive taking his fight to LLMs when search engines and OSs have been leeching and hijacking content for ages.
There was never a guarantee -- for anyone in any industry at all -- that what worked in the past will always continue to work. That is a regressive attitude.
However I do have concerns about Google and other monopolies replacing large swaths of people who make their livings doing things that can now be automated. I am not against automation but I don't think the disruption of our entire societal structure and economy should be in the hands of the sociopaths that run these companies. I expect regulation to come into play once the shit hits the fan for more people.
But why did you want the traffic? Was your revenue from ad impressions, or were you perhaps being paid by the city of Barcelona to provide useful information to tourists? If the former, I get that this hurt you. If the latter, was this a failure or a success?
Parent has the right question here: why did you want the traffic? Did you intend for anything good to happen to those people?. I'm going to guess not; there's hardly a scenario where people who complain about loss traffic and mean that traffic any good.
Except it kind of does. Almost all high-quality free content on the Internet has been made by hobbyists just for the sake of doing it, or as some kind of expense (marketing budget, government spending). The free content is not supposed to make money. An honest way of making money with content is putting up a paywall. Monetizing free content creates a conflict of interest, as optimizing value to publisher pulls it in opposite direction than optimizing for value to consumer. Can't save both masters, and all. That's why it's effectively a bullet-proof heuristic, that the more monetization you see on some free content, the more wrong and more shit it is.
Put another way, monetizing the audience is the hallmark of slop.
The hypocrisy and self indulgent fickleness of some educated idiots on HN truly is something else at times.
They just prefer internet users to consume their ads, rather than the ads of the content creators.
Thank $deity that there are still search engines you can just pay for. A honest exchange of money for value is getting quite uncommon these days.
Of course! It's certainly better to ruin the few sites that support their attempts at high quality content with ad revenue. Much better to let Google have that money, because of course the tech giant has nothing to do with enhsittifying everything through ad revenue of its own and pervasive tracking, or enabling ever worse content through SEO and AI gaming.
You can appreciate that a modest site trying to survive through ads isn't necessarily evil just because it looks for a way to make money off its content?
I mean, what specific harm are you referring to? Particularly compared to the much more obvious harm of Google absorbing ever more of the web in favor of its tentacled surveillance/SEO gaming machine.
No, I cannot appreciate that.
You expect that people should be obligated to conduct their efforts at creating readable information for free, unless they want your moral disdain?
Particularly laughable notions from someone enjoying a site deeply embedded in the ad-funded Silicon Valley parasitic consumer surveillance landscape.
It's not necessarily evil, just statistically very likely so :). It's still affected by the conflict of interest, though. Making money off content directly means you either ask readers to pay up, or you extract that payment somehow, whether they want it or not. And since the site isn't asking...
> I mean, what specific harm are you referring to? Particularly compared to the much more obvious harm of Google absorbing ever more of the web in favor of its tentacled surveillance/SEO gaming machine.
At scale, http://jacek.zlydach.pl/blog/2019-07-31-ads-as-cancer.html.
At individual interaction level, think of it as smoking. One cigarette isn't going to kill you. Hell, some smoking might even lose your weight! But it still affects your behavior short-term in a self-reinforcing way, and long-term, it's gonna ruin your health. A site monetizing content with ads is like a store or library that lets you read for free, if you take a whiff or three of the specific brand of cigarettes they're sponsored by. A couple interactions may not hurt, but continuous exposure definitely will.
Just because the damage happens to your brain instead of your lungs and immune system, doesn't mean it's OK now. It's still an asshole move to expose your fellow humans to poison.
And finding a means of funding for content is not bad thing even if it involves a conflict of interest. You're painting it as if it were some sort of nefarious activity when in reality it consists of "here's our content, much of it is authentic, verifiably useful (people want it after all and keep reading) and also, we make money off these very visible ads right here. If anything, it's a better model than internally recommending things in content in a dishonest way.
What's more, compared to instead handing that content and those eyeballs over to Google, the top monster itself of online ads, dark patterns and gamed suggestions, it's the much better option.
Your underlying narrative seems to be that people trying to use their efforts at content online to make money is somehow inherently morally wrong, and that's absurd. It's particularly ridiculous when, as in this case, the alternative is a colossal advertising/tech corporation essentially stealing that content to suck away views from these much tinier sites.
We sold experiences, thus we created a lot of free content from local experts and hoped that they would buy some of the tickets through our website.
As much as I dislike Google, people who create content FOR google are infinitely worse IMO as they bury all the genuine content created by people without a profit motive. You can always go find a business model that doesn't depend on Google driving traffic to your website.
They make money by you reading their stuff, not by you actually spending money in the place.
It's a legitimate complaint, and it sucks for your business. But I think this demonstrates that the sort of quality content you were producing doesn't actually have much value.
They don't present it all, they summarize it.
And let's be serious here, I was being polite because I don't know the OPs business. But 99% of this sort of content is SEO trash and contributes to the wasteland that the internet is becoming. Feel free to point me to the good stuff.
Like to society the value is in the hotel, everything else is just fluff around it that never had any real value to begin with.
> Feel free to point me to the good stuff.
Travel bloggers and vloggers, but that is an entirely different unaffected industry (entertainment/infotainment).
I've no doubt some good ones exist, but my instinct is to ignore every word this industry says because it's paid placement and our world is run by advertisers.
This will be quite interesting in the future. One can usually tell if a blog post is stale, or whether it’s still relevant to the subject it’s presenting. But with LLMs they’ll just aggregate and regurgitate as if it was a timeless fact.
It sucks that Google created the problem on both ends, but the content OP is referring to costs way more to produce than it adds value to the world because it has to be padded out to show up in search. Then Google comes along and extracts the actual answer that the page is built around and the user skips both the padding and the site as a whole.
Google is terrible, the attention economy that Google created is terrible. This was all true before LLMs and tools like Perplexity are a reaction to the terrible content world that Google created.
If Google preferred websites that cut the fluff, then website operators would have an incentive to make useful websites, and Google wouldn't have as much of an incentive to provide the answer in a snippet, and everyone wins.
I guess it's hard to rank website quality, so Google just prefers verbose websites.
Google has at least two incentives to provide that answer, both of which wouldn't change. The bad one: they want to keep you on their page too, for usual bullshit attention economy reasons. The good one: users prefer the snippets too.
The user searching for information usually isn't there to marvel at beauty of random websites hiding that information in piles of noise surrounded by ads. They don't care about websites in the first place. They want an answer to the question, so they can get on with whatever it is they're doing. When Google can give them an answer, and this stops them from going from SERP to any website, then that's just few seconds or minutes of life that user doesn't have to waste. Lifespans are finite.
The only reason that users prefer snippets is because websites hide the info you are looking for. The problem is that the top ranked search results are ad-infested SEO crap.
If the top ranked website were actually designed with the user in mind, they would not hide the important info. They would present the most important info at the top, and contain additional details below. They would offer the user exactly what they want immediately, and provide further details that the user can read if they want to.
Think of a well written wikipedia article. The summary is probably all that you need, but it's good that the rest of the article with all the detail is there as well. I'm pretty sure that most people prefer a well designed user-centric article to the stupid Google snippet that may or may not answer the question you asked.
Most people looking for info don't look for just a single answer. Often, the answer leads to the next question, or if the answer is surprising, you might want to check out if the source looks credible, etc. Even ads would be helpful, if they were actually relevant (eg. if I am looking for low profile graphic cards, I'd appreciate an ad for a local retailer that has them in stock).
But the problem is that website operators (and Google) just want to distract you, capture your attention, and get you to click on completely irrelevant bullshit, because that is more profitable than actually helping you.
Yes, Google hijacked images for some time. But in general there has "always" been the option to tell Google not to display summaries etc with these meta tags:
<meta name="googlebot" content="noarchive">
<meta name="googlebot" content="nosnippet">If I visit your site from Google with my browser configured to go straight to Reader Mode whenever possible, is my visit more useful to you than a summary and a link to your site provided by Perplexity? Why does it matter so much that visitors be directly on your content?
* User asks proprietary web browser to fetch content and render it a specific way, which it does
* User asks proprietary web service to fetch content and render it a specific way, which it does
The technical distinction is that there's a network involved in the second scenario. What is the moral distinction?
If Perplexity visit the site once and cache some info to give to multiple users, that is stealing traffic numbers for ad value, but also taking away the ability from the site owner to get realistic ideas of how many people are using the information on their site.
Additionally, this is AI we are talking about. Whos to say that the genrated summary of information is actually correct? The only way to confirm that, or to get the correct information in the first place, is to read the original site yourself.
As someone who uses Perplexity, I often do do this. And I don't think I'm particularly in the minority with this. I think their UI encourages it.
I wonder if it would be slightly less sad if we all had our own decentralized crawlers that simply functioned as extensions of ourselves.
This is something I'm (slowly) working on myself. I have a local language model server and 30 tb usable storage ready to go, just working on the software :)
I have another comment that says something similar, but: is valuing a website based on basic traffic still a thing? Feels very 2002. It's not my wheelhouse, but if I happened to be involved in a transaction, raw traffic numbers wouldn't hold much sway.
I've never used that tool (and don't plan to) so I don't know. If they just embed the content in an iframe or something then there's no issue (but then there's no need or point in scraping). If they're just scraping to train then I think you also imply there's no issue. If they're just copying your content (even if the prompt is "Hey Perplexity, summarise this article <ARTICLE_TEXT>") then that's vanilla infringement, whether they lie about their UA or not.
Reality is, the law doesn't matter if you're big enough. As long as they're not stealing content from the big ones, they're going to be fine.
I'm seriously sick of that whole "laundering copyright via AI"-grift - and the destruction of the creative industry is already pretty noticable. All the creatives who brought us all those wonderful masterworks with lots of thought and talent behind, they're all going bankrupt and getting fired right now.
It's truly a tragedy - the loss of art is so much more serious than people seem to think it is, considering how integral all kinds of creative works are to a modern human live. Just imagine all of that being without any thought, just statistically optimized for enjoyment... ugh.
Can you explain what you mean by this? I’d be interested to know what jobs have been lost to AI (or if you are talking about something else)
The most extreme situation is concept artists right now. Essentially, the entire profession has lost their jobs in the last year. Or casual artists making drawings for commission - they can't compete with AI and mostly had to stop selling their art. Similar is happening to professional translators - with AI, the translations are close enough to native that nobody needs them anymore.
The book market is getting flooded with AI-crap, so is of course the web. Authors are losing their jobs.
Currently, it seems to be creeping into the music market - not sure if people are going to notice/accept AI-made music. All the fantastic artists creating dubs are starting to go away as well, after all you can just synthesize their voices now.
It's quite sad, all considered.
Except, it can't possibly be like that - that would kill the Internet as you know it. It makes sense to consider scrapping for purposes of training as infringement - I personally disagree, I'm totally on the side of AI companies on this one, but there's a reasonable argument there. But in terms of me requesting a summary, and the AI tool doing it server-side before sending it to me, without also adding it to the pile of its own training data? Banning that would mean banning all user-generated content websites, all web viewing or editing tools, web preview tools, optimizing proxies, malware scanners, corporate proxies, hell, maybe even desktop viewers and editing tools.
There are always multiple programs between your website and your user's eyeballs. Most of them do some transformations. Most of them are third-party, usually commercial software. That's how everything works. Software made by "AI company" isn't special here. Trying to make it otherwise is some really weird form of prejudice-driven discrimination.
It's not transformative, it uses the entire work, the work copied is not a summary of facts, and insofar as their is a market at all, it circumvents that market. It fails every test.
(training OTOH is inherently transformative, and I suspect likely to turn out to be a fair use)
I would say though, it feels like the training argument may ultimately lead to a similar outcome, though it’s a bit more ideological and less tangible than regurgitating the results of a query. Services like chatgpt are already being used a google replacement by many people, so long term it may reduce clicks from search as well.
Perplexity has source references. I find myself visiting the source references. Especially to validate the LLM output. And to learn more about the subject. Perplexity uses a Google search API to generate the reference links. I think a better strategy is to treat this as a new channel to receive visitors.
The browsing experience should be improved. Mozilla had a pilot called Context Graph. Perhaps Context Graph should be revisited?
> In a world where users ask Perplexity these Help questions about my SaaS, Perplexity may answer them and I would lose all the insights because I never get any traffic.
This seems like a missing feature for analytics products & the LLMs/RAGs. I don't think searching via an LLM/RAG is going away. It's too effective for the end user. We have to learn to work with it the best we can.
Alternative take: Perplexity is protecting users' privacy by not exposing them to be turned into "insights" by the SaaS.
My general impression is that the subset of complaints discussed in this thread and in the article, boils down to a simple conflict of interest: information supplier wants to exploit the visitor through advertising, upsells, and other time/sanity-wasting things; for that, they need to have the visitor on their site. Meanwhile, the visitors want just the information without the surveillance, advertising and other attention economy dark/abuse patterns.
The content is the bait, and ad-blockers, Google's instant results, and Perplexity, are pulling that bait off the hook for the fish to eat. No surprise fishermen are unhappy. But, as a fish, I find it hard to sympathize.
This appears to be self-contradictory. If you let an LLM to be trained* on “all the books” (posts, articles, etc.) in the world, the implication is that your potential readers will now simply ask that LLM. Not only will they pay Microsoft for that privilege, while you would get zilch, but you would not even know they ever read the fruits of your research.
* Incidentally, thinking of information acquisition by an ML model as if it was similar to human reading is a problematic fallacy.
Just to clarify, Perplexity is not spoofing a user agent, they're legitimately using a headless Chrome to fetch the page.
The author just misunderstood their docs [0]: when they say that "you can identify our web crawler by its user agent", they're talking about the crawler, not the browser they use for ad hoc requests. As you note, crawling is different.
Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) HeadlessChrome/119.0.0.0 Safari/537.36
They are spoofing it to pretend to be a desktop Chrome one: Mozilla/5.0 (Windows NT 10.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/111.0.0.0 Safari/537.36I don't personally have a problem with spoofing user agents, but yeah, they're either spoofing or for some reason they're truly using a non-headless Chrome.
No, easier to just ask a simple question: Does the company respect the access rules communicated via a web standard? No? In that case hard deny access to that company.
These companies don't need to be given an inch.
So should Firefox not allow changing the user agent in order to bypass websites that erroneously claim to not work on Firefox?
Ad block isn’t the same problem because it doesn’t and can’t steal the creator’s data.
Arguably it does. That topic has been debated endlessly and there are plenty of people on HN who are willing to fiercely argue that adblock is theft.
I happen to agree with you that adblock doesn't steal data, but I'm also completely unsure why interacting with a tool over a network suddenly turns what would be acceptable on my local computer into theft.
Website owners decide to stop publishing because it’s not rewarded by a real human visit anymore?
Then perplexity and the like won’t have new information to train their models on and no sites to answer the questions.
I think there is a real content dilemma here at work. The incentives of Google and website owners were more or less aligned.
This is not the case with perplexity.
Is it necessary to load the JavaScript for it to count as a visit? What if I access the site with noscript?
Or is it only a visit if I see all your recommended content? I usually block those recommendations so that I don't get distracted from the article I actually came to read—is my visit a less legitimate visit than other people's?
What exactly is Perplexity doing here that isn't okay that people don't already do with their local user agents?
At the very least they get exposed to your website name.
Notice your product/service if you get lucky.
Become a customer at a later visit.
We are talking about cutting the first step off so that everything which may come afterwards is cut off as well.
[0] Which, incidentally, is entirely normal. robots.txt is for the web crawler that indexes, not for ad hoc requests.
I'm more than happy to see all the websites with attention economy business models to shut down. Yes, that might be 90% of the Internet. That would be the 90% that is poisonous shit.
The attention economy will not die. Because it’s hasn’t for the last 100 years. The profits just shift to where the attention is now.
Whats stopping perplexity caching this info say for 24 hours, and then redisplaying it to the next few hundred people who request it?
Does this place HN's rampant use of archive.md on the same moral footing as Perplexity?
It's in the title of TFA: they're being dishonest about who they are. PerplexityBot seems to understand that robots.txt is addressed to it.
It's understood that site operators have a right to use the User-Agent to discriminate among visitors; that's why robots.txt is a standard. Crawlers that disrespect the standard have for many years been considered beyond the pale; thieves and snoopers. TFA's complaint is entirely justified.
First, I'm ignoring the output of Perplexity. I have no reason to believe that they gave the LLM any knowledge about its internal operations, it's just riffing off of what OP is saying.
Second, PerplexityBot is the user agent that they use when crawling and indexing. They never claimed to use that user agent for ad hoc HTTP requests (which are notably not the same as crawling).
Third, I disagree that anyone has an obligation to be honest in their User-Agent. Have you ever looked at Chrome's user agent? They're spoofing just about everyone, as is every browser. Crawlers should respect robots.txt, but I'd be totally content if we just got rid of the User-Agent string entirely.
Is that a distinction without a difference?
I think the robots.txt RFC was addressed specifically to crawlers; so technically "ad hoc" requests generated automatically (i.e. by robots) aren't included. But the distinction operators would like to make is between humans and automata. Whether some automaton is a crawler or not isn't relevant.
* The pattern of requests can be very problematic. Impolite crawlers are totally capable of taking down a website by hitting it over and over and over again for hours in a way that humans won't.
* Crawlers are generally used to build search indexes, so instructing them about URLs that would be inappropriate to have show up in a search is relevant.
The behavior that OP is complaining about is that when the user pastes a URL into Perplexity, Perplexity fetches that URL. Neither the traffic pattern nor the persistence profile are remotely similar to typical crawler behavior. As far as I can see there's almost nothing to distinguish it from someone using Edge and then using Edge's built-in summarizer.
And if we took the analogy to the other end, one could argue that all crawlers have to be kicked off manually at some point...
The problem is here in reality the differentiation is somewhat more understood.
The honor system web is going away, that's for sure.
There are a lot of people making this assumption about the way Perplexity is working, but there is no evidence in TFA that Perplexity is caching its ad hoc requests.
And even if they were, what's left unsaid is why it even would matter if 500 views turned into one. It matters either because of lost ad revenue or lost ability to track the users' behavior. Personally, I'm okay with moving past that phase of the internet's life and look forward to new business models that aren't built around getting large numbers of "views".
So, a caching proxy? That has its own issues, but it's the opposite of access by automaton. One button press causes less than one access to the server. Though one button press still results in one user view, so it's only reducing loads in some ways.
But also is that happening here?
> And if we took the analogy to the other end, one could argue that all crawlers have to be kicked off manually at some point...
One button press causing a million page loads is access by automaton. The distinction seems pretty simple to me.
It's not really a dilemma.
This is exactly what copyright serves to protect authors from. Perplexity copied the content, and in doing so directly competes with the original work, destroying it's market value and driving the original author out of business. Literally what copyright was invented to prevent.
It's the exact same situation as journalists going after Google & social media embeds of articles, which these sites propagandized as "prohibiting hyperlinking", but the issue has been the embedded (summary of the) content. Which people don't click through, and this is the entire point of those features for platforms like Facebook; Keeping users on facebook and not leaving.
This is why quite a few jurisdictions agreed with the journalists and moved to institute restrictions on such embedding.
By all practical considerations, perplexity is doing the exact same thing and trying to deflect with "we used an AI to paraphrase".
> The incentives of Google and website owners were more or less aligned.
The key difference here is that linking is and always has been fine. Google's Book search feature is fair use because the purpose is to send you to the book you searched for, not substitute the book.
Google's current AI summary feature is effectively the same as Perplexity. People don't click through to the original site, the original site doesn't get ad impressions or other revenue, and is driven out of business.
> What will happen if:
What will happen is what already is happening: Journalists are driven out of business, replaced by AI slop.
And then what? AI needs humans creating original content, especially for things like journalism and fact-finding. It'd be an eternal AI winter, all LLMs doomed to be stuck in 2025.
It's in every AI developer's best interest to halt the likes of Perplexity immediately before they irreparably damage the field of AI.
And really, "I'm going to replace my entire news intake with the AI slop even if it's entirely hallucinated lies or propaganda" is perhaps not something you ought to say out loud.
You're missing something crucial. Yes, there may be hallucinations, but that's why there's preference for AI that provides citations; they can be easily checked. And also, in my experience, the summaries are usually decent; the sources which tend to yield pretty broken summaries also tend to intersperse unrelated material (ads, previews to other things, etc) in the content, or they're doing wild transforms with JS for example to create unnecessary eye candy. And coincidentally, I'd rather avoid those sources where possible, so it becomes an even greater win for me as user to have AI prefilter said content. Thus the sources getting those precious views in the end become the ones that respect their users' time and aversion to distractions.
It's in every AI developer's best interest to halt the likes of Perplexity immediately before they irreparably damage the field of AI.
That’s exactly the problem and we all know that it will happen.
I guess if you're doing it for a living sure, but most content I consume online is created without incentive (social media, blogs, stack overflow).
I write a fair amount and have been for a few years. I like to play with ideas. If an llm learned from my writing and it helped me propagate my ideas, I'd be happy. I lose on social status imaginary internet points but I honestly don't care much for them.
The craziest one is the stack overflow contributors. They write answers for free to help people become better programmers but they're mad an llm will read their suggestions and answer questions that help people become better programmers. I guess they do it for the glory of having their handle next to the answer?
Yes, it's hardly surprising that people find upvotes and direct social rewards more exciting than being slurped somewhere into GPT-4's weights.
It's not like they're getting validation from most readers anyway. People who vote and comment on answers are playing the SO social/karma game and will continue to do so whether GPT-4 exists or not. Conversely, people who'll find answers via an LLM instead of viewing it on SO are people who wouldn't bother logging in to SO, even if they had accounts on it in the first place.
People are complaining about losing the audience they never had.
If the issue here was "just" training LLMs, like some AI bros want to deflect it to be, the conversation around this topic would be very different, and I would be enthusiastically defending the model trainers.
But that's not this conversation. These are companies that are trying to fold our permissively-license content into weights, close source it, and make themselves the only access point, all while pre-emptively perform regulatory capture with all the right DEI buzzwords so that the open source variants are sufficiently demonized as "alt-right" and "dangerous".
The thing that truly frightens me is that (even here on Hacker News) there is an increasing number of people that have fallen for the DEI FUD and are honestly cheering on the Sam Altmans of the world to control the flow of information.
Because you're not getting the ad impressions anymore. The harsh reality is that people do not click on to sources, so when sites like Perplexity copy your content, you lose the revenue on that content.
This, in turn, drives all real journalism out of business. And then everyone's screwed, including these AI reposting sites.
For what it's worth, the Stack Exchange terms of service do prohibit AI generated content. I'm not sure how they actually enforce that, and in practice as the LLMs improve it's going to be almost impossible to reliably detect.
It is not simply "nice", or for internet points, to take time to answer other people's questions.
Being able to pass on knowledge is the glue of society and civilization. Cynicism about the value or reason of doing so is not a replacement for a functioning structure to educate people who want to learn or to point them in the right direction.
In my experience they do it for points and kudos. Having people get your answers from LLMs instead of your answer on SO stops people from engaging with the gamification tools and so users get less points on the site.
For technical content of value to professionals, much of that is hosted by vendors or industry organizations. Those tend to get their revenue in other ways and don't care about companies scraping their content for AI model training. Like the IETF isn't going to stop publishing new RFCs just because Perplexity uses them.
I’m not saying you’re wrong, but why? And what do you mean by “your data” here?
The website that they created.
This feels like the fundamental core component of what copyright allows you to forbid.
> Everything from ad blockers to reader mode to screen readers do exactly the same thing that Perplexity is doing here, with the only difference being that they tend to be exclusively local
Which is a huge difference. The latter is someone asking for a copy of my content (from someone with a valid license, myself), and manipulating it to display it (not creating new copies, broadly speaking allowed by copyright). The former adds in the criminal step of "and redistributing (modified, but that doesn't matter) versions of it to users without permission".
I mean, I'm all for getting rid of copyright, but I also know that's an incredibly unpopular position to take, and I don't see how this isn't just copyright infringement if you aren't advocating for repealing copyright law all together.
I'm assuming that if I write code by hand for every part of the TCP/IP and HTTP stack I'm safe.
What if I use libraries written by other people for the TCP/IP and HTTP part?
What if I use a whole FOSS web browser?
What about a paid local web browser?
What if I run a script that I wrote on a cloud server?
What if I then allow other people to download and use that script on their own cloud servers?
What if I decide to offer that script as a service for free to friends and family, who can use my cloud server?
What if I offer it for free to the general public?
What if I start accepting money for that service, but I guarantee that only the one person who asked for the site sees the output?
Can you help me to understand where exactly I crossed the line?
> What if I decide to offer that script as a service for free to friends and family, who can use my cloud server?
You're allowed to make copies and adaptations in order to utilize the program (website), which probably covers a cloud server you yourself are controlling. You aren't allowed to do other things with those copies though, like distribute them to other people.
Payment only matters if we're getting into "free use" arguments, and I don't think any really apply here.
I think you're probably already in trouble with just offering it to family and friends, but if you take the next step offering it to the public that adds more issues because the copyright act includes definitions like "To perform or display a work “publicly” means (1) to perform or display it at a place open to the public or at any place where a substantial number of persons outside of a normal circle of a family and its social acquaintances is gathered; or (2) to transmit or otherwise communicate a performance or display of the work to a place specified by clause (1) or to the public, by means of any device or process, whether the members of the public capable of receiving the performance or display receive it in the same place or in separate places and at the same time or at different times."
No one is distributing copies of anything to anyone then apart from the website that owns the content lawfully distributing a copy to the user.
Also why is a paid web browser any different than a free one?
I'm unfamiliar with the legal questions, but in 2024 I have a very hard time seeing an ethical distinction between running some proprietary code on my machine to complete a task and running some proprietary code on a cloud server to complete a task. In both cases it's just me asking someone else's code to fetch data for my use.
> I have a very hard time seeing an ethical distinction between running some proprietary code on my machine to complete a task and running some proprietary code on a cloud server to complete a task
It's important to recognize that copyright is entirely artificial. Congress went "let's grant creators some monopolies on their work so that they can make money off of it", and then made up some arbitrary lines for what they did and did not have a monopoly over. There's no principled ethical distinction between what is on one side of the line and the other, it's just where congress drew the arbitrary line in the sand. It then (arguably) becomes unethical to do things on the illegal side of the line precisely because we as a society agreed to respect the laws that put them on the illegal side of the line so that creators can make money in a fair and level playing field.
Sometimes the lines in the sand were in fact quite problematic. Like the fact that the original phrasing meant that running a computer program would almost certainly violate that law. So whenever that comes up congress amends the exact details of the line... in the US in the case of computers carving out an exception in section 117 of the copyright act. It provides that (in part)
> it is not an infringement for the owner of a copy of a computer program to make or authorize the making of another copy or adaptation of that computer program provided:
> (1) that such a new copy or adaptation is created as an essential step in the utilization of the computer program in conjunction with a machine and that it is used in no other manner
and provides the restriction that
> Adaptations so prepared may be transferred only with the authorization of the copyright owner.
By my very much not a lawyer reading of the law, those are the relevant parts of the law, they allow things like local ad-blockers, they disallow a third party website which downloads content (acquiring ownership on a lawfully made copy), modifies it (valid under the first exception if that was a step in using the website) and distributes the adapted website to their users (illegal without permission).
Your logic would also imply that viewing a website through a VPN not operated by yourself would require the VPN operator to have a redistribution license for all the content on the website which is not the case.
How do you think google is able to scrape whatever they like and redistribute summaries of the pages they have visited without consulting everyone who has ever made a website for a redistribution license.
That being said, Copyright is not enforced or interpreted consistently. It seems that individual cases can be decided based on what people ate for lunch on the day of the case, who the litigants are, and maybe the alignment of the planets.
The key, as many here have missed, is authentication and authorization. You may have authorization to log in and view movies on Netflix. Not to rebroadcast them. Even the question of a VCR for personal use was debated in the past.
Distributing your own scripts and software to process data is not the same as distributing arbitrary data those scripts encountered on the internet for which you don’t have a license.
If someone wrote an article, your reader transforms it based on your authenticated request, and your user would have an authorized subscription.
But if that reader then sent the article down to a remote server to be processed for distribution to unlimited numbers of people, it would be “pirating” that information.
The problem is that much of the Web is not properly guarded against this. Xanadu had ideas about micropayments 30 years ago. Take a look at what I am building using the current web: https://qbix.com/ecosystem
LEGAL ANALYSIS
Much of the content published on the Web isn’t secured with subscriptions and micropayments, which is why the whole thing becomes a legal battle as silly as “exceeding authorized access” which landed someone like Aaron Swartz in jail.
In other words, it is the question of “piracy”, which has acquired a new character only in that the AI is trained on your data and transforms it before it republishes it.
There was also a lawsuit aboot scraping LinkedIn, which was settled as follows: https://natlawreview.com/article/hiq-and-linkedin-reach-prop...
Legally, you can grant access to people subject to a certain license (eg Creative Commons Share Alike) and then any derived content must have its weights opened. Similar to, say, Affero GPL license for derivative software.
Both are, the difference is that your browser doesn't transfer the copy to a new legal entity after modifying it. Rather the browser is under the control of the end user and the end user owns the data (not the copyright, but the actual instance of the data) the whole time.
> Your logic would also imply that viewing a website through a VPN not operated by yourself would require the VPN operator to have a redistribution license for all the content on the website which is not the case.
It doesn't because the VPN doesn't modify it, and the law explicitly distinguishes between the two cases and allows for transferring in the case of exact copies (provided you transfer all rights). I left this part of section 117 out because it wasn't relevant, but I'll quote it here
> Any exact copies prepared in accordance with the provisions of this section may be leased, sold, or otherwise transferred, along with the copy from which such copies were prepared, only as part of the lease, sale, or other transfer of all rights in the program. [And then the portion of the paragraph I quoted above] Adaptations so prepared may be transferred only with the authorization of the copyright owner.
> How do you think google is able to scrape whatever they like and redistribute summaries of the pages they have visited without consulting everyone who has ever made a website for a redistribution license.
A fair use argument, which I think is less likely (and I'd go so far as to say unlikely) to apply to a service like perplexity.ai but is ultimately a judgement call that will be made by the legal system and like all fair use arguments has no clear boundaries.
* User types an address into Perplexity
* Perplexity fetches the page, transforms it, and renders some part of it for the user
But this flow is not:
* User types an address into Orion Browser
* Orion Browser fetches the page, transforms it, and renders some part of it for the user
Regardless of the legal question (which I'm also skeptical of), I'm especially unconvinced that there's a moral distinction between a web service that transforms copyrighted works in an ad hoc manner upon a user's specific request and renders them for that specific user vs an installed application that does exactly the same thing.
Also, the author of TFA has already gotten themselves deindexed, the behavior they're complaining about now is that if someone copies and pastes a link into Perplexity it will go fetch the page for the user and summarize it.
This scenario presupposes that the user has a link to a specific page. I suspect that in nearly all cases that link will be copied from the address bar of an open tab. This means that most of the time the site will actually get double the traffic: one hit when the user opens it in the browser and a second when Perplexity asks for the page to summarize it.
However, please see the Aereo case [0] for a possibly analogous case. I am allowed to have a DVR. There is no law preventing me from accessing my DVR over a network. Or possibly even colocating it in a local data center. But Aereo definitely crossed a line. Also see Vidangel [1]. The fact that something is legal to do at home, does not mean that I can offer it as a cloud service.
I expect you're right. Although Perplexity thinks they're well within the law[0]. Are they correct? I guess we'll see....
[0] https://www.perplexity.ai/search/Why-are-you-2wJteqZ4SUCqPjk...
1. Asking for a copy of your content
2. Manipulating the content
3. Redistributing the content to the end-user who requested it
Ditto for the LLM that has been asked by the end user to fetch your content and show it to them (possibly with a manipulation step e.g. summarization).
I don't think there's a legal, copyright distinction between doing that on a server vs doing that on a local machine. And, for example, if there were a difference: using a browser on a remote desktop would be illegal, or using curl on a machine you were SSHed into would be illegal. Also, an LLM running locally on your machine (doing the exact same thing) would be legal!
I understand that it's inconvenient and difficult to monetize content when an LLM is summarizing it, and hard to upsell other pages on a website to users when they aren't coming to your website and are instead accessing it through an LLM. But legally I think there's not an obvious distinction on copyright grounds, and if there were (other than a very fine-grained ban on specifically LLMs accessing websites, without any general principle behind it), it would catch up a lot of legitimate behavior in the dragnet.
I'd also point out that in the U.S., search engines have passed the "Fair Use" test of exemption from copyright — I think it would be very hard to make a distinction between what a search engine is doing (which is on a server!) and what an LLM is doing based on trying to say copyright distinguishes between server vs client architectures.
If you're in control of the server there's a plausible argument that you aren't transferring it. When perplexity is in control of the server... I don't see it. A traditional ad-blocker isn't "redistributing the content to the end-user who requested it" because it's the end user who has ownership over the data the whole time (note: not the copyright, the actual individual instance of the data). Unlike with a server run by a third party there is no third party legal entity who ever has the data.
You could conceivably make "ublock origin except it's a proxy run by a third party and we modify the website on the proxy", I'd agree that that has the same problem as a service like perplexity (though a different fair use analysis and I'm not sure what way that would go).
> I'd also point out that in the U.S., search engines have passed the "Fair Use" test of exemption from copyright — I think it would be very hard to make a distinction between what a search engine is doing (which is on a server!) and what an LLM is doing based on trying to say copyright distinguishes between server vs client architectures.
Well, sure. It's easy to distinguish between an LLM summarizing content and a traditional search engine though (and in ways relevant to the fair use analysis), just not based on the server client architecture.
Disclaimer: Not a lawyer, not legal advice, and so on.
(Although it is relevant in disproving your point: I can pay an archival service to back up data I legally have the right to view, even if the backup is then on their server, and despite the service being a different legal entity than me. And they can give me a copy of it later, too.)
> You could conceivably make "ublock origin except it's a proxy run by a third party and we modify the website on the proxy", I'd agree that that has the same problem as a service like perplexity (though a different fair use analysis and I'm not sure what way that would go).
So, running a local LLM version of Perplexity that does exactly the same thing is legal, but Perplexity is illegal, because "a third party legal entity has the data"?
It seems like some website owners want to have their cake and eat it too. They want their content indexed by Google and other crawlers in order to drive search traffic but they don't want their content used to train AI models that benefit other companies. At some point they're going to have to make a choice.
Parsing, rendering, and trusting that the payload is consistent from request to request is your problem. You can connect to my server, or not. I really don’t care. What you cannot do is dictate how my server responds to your request.
If I think a given request is from an LLM training module, I don't have any legal obligation whatsoever to return my original content. Or a 400-series response. If I want to intersperse a paragraph from Don Quixote between every second sentence, that's my call.
The client is under no obligation to be truthful in its communications with a server. Spoofing a User-Agent doesn't "dictate" anything. Your server dictates how it responds all on its own when it discriminates against some User-Agents.
Even redistributing secret material you found on an accidentally open S3 bucket, without spoofing UA, could be considered intrusion if it was obvious the material was intended to be secret and you acted with bad intent.
It is also a technical fact that a client can send any header it wants.
Accessibility can mean everything from a blind person wanting to interacting with a website using voice, to someone recovering from a surgery and wanting something to reduce unnecessary popups and clicks on a website to get to the information they need. Accessibility is in the eye of the accessor, and AI is what enables them to achieve it.
The way I see it, AI is not a robot and doesn't need to look at robots.txt. Rather, AI is my low-cost secretary.
I don't think you are seeing it very clearly then. Your secretary can also be a robot. What do you think an AI is if not a robot??
It doesn't "need" to look at robots.txt because nothing does.
I'm okay with this world, as a tradeoff. I'm not sure users should have _the right_ to reformat others' content.
N O T H I N G.
Could you change the way printed news magazines showed their content? No. Then, why is that a problem?
Btw nobody clicks on sources. NOBODY.
I always click on sources to verify what an LLM in this case says. I also hear the claim that a lot about people not reading sources (before LLM it was video content with references) but I always visited the sources. Is there a statistics or studies that actually support this claim? Or is it just a personal experience, of people (including me) enforcing it as generic behavior of all people?
Also perplexity shows the videos on their site, you cannot go to youtube, you have to start it on their site, and then you have to click on the youtube player's logo in the lower right to get to the site.
Perplexity is getting greedy.
What distinguishes these two situations?
* User asks proprietary web browser to fetch content and render it a specific way, which it does
* User asks proprietary web service to fetch content and render it a specific way, which it does
The technical distinction is that there's a network involved in the second scenario. What is the moral distinction?
Why is it that a proprietary web service manipulating content on behalf of a user is "publishing" content illegally, while a proprietary web browser doing the exact same kind of transformations is not? Assume that in both cases the proprietary software fetches the data upon request, does not cache it, and does not make the transformed content available to other users.
> * User asks proprietary web service to fetch content and render it a specific way, which it does
That sounds like Google Translate to me, when pasting a URL.
Bonus points if instead of pasting a URL directly, it is submitted to one of the Internet Archive-like sites; and then submit that archive URL to Google Translate. That would be download and adaptation (by Google Translate) of the download and adaptation[1] (by Internet Archive) of the original content.
[1]: These archive sites usually present the content in a slightly different way. Granted, it's usually just adding stuff around the page, e.g. to let you move around different snapshots, but that's still showing stuff that was originally not there.
If the company wants a copy of the files for your own use, then that is a bit different. When accessing large number of files at once, robots.txt is useful to block it. If they can get a copy of the files in a different way (assuming the files are intended to be public anyways), then they might do so. However, even in this case, still they should not use unethical business practices, false advertising, etc; and, they should also avoid user-agent spoofing.
It is transforming the content for you, an authorized party.
That is not the same as then making derivative copies and distributing the information to others without paying. For example, if I bought a ticket to a show, taped it and then distributed it to everyone, disregarding that the show prohibited this.
If I shared my Netflix password with up to 5 others, at least I can argue that they are part of my “family” or something. But to unlimited numbers of people? Why would they pay for netflix, and how would the shows get made?
I am not necessarily endorsing government force enforcing copyright, which is why I have been building a solution to enforce it at the tech level: https://Qbix.com/ecosystem
However, what you say about credentials is still valid in the case of private data; this is why you should run the program locally and not use some other company's remote service for this use. (Well, it is one reason why. Other reason is all of the other bad stuff they do with the service.)
It is also valid about credentials, even if it is published but requires a password to access using that service; but even then, if you would ignore copyright, you can just use a different copy of the same file (which you might make by yourself).
None of this is meaning that you cannot pay for it, if they accept payment. It is also not meaning that whoever made it is required to give it away for free. What it is meaning, is that if you have a copy, you do not have to worry about copyright and other legal mess; you can just to do it; a license is not required.
However, it is also another issue how much power big companies are wasting with your data, whether they are authorized to access it or not. This is potentially a reason to disallow some uses, but that is independent from copyright (which is bad, anyways).
If what Perplexity is doing is illegal, is it illegal to run an open-source LLM on your own machine, and have it do the same thing? If so, how are ad blockers or Reader Modes or screen readers legal?
And if it's legal to run an open-source LLM on your own machine, is it legal to run an open-source LLM on a rented server (e.g. because you need more GPUs)? And if that's legal, why is it illegal to run a closed-source LLM on servers? Could Perplexity simply release the model weights and keep doing what they're doing?
1) a user-agent which makes an authenticated and authorized request for data, and delivers to the user
2) a user who then turns around and distributes the data or its derivatives to users in an unauthorized manner
A “dumber” example would be whether I can indefinitely cache and index most of information via the Google Places API, as long as my users request each item at least once. Can I duplicate all that map or streetview photo information that google paid cars to go around and photograph? Or how about the info that Google users entered as user-generated content?
THE REQUIREMENT TO OPEN SOURCE WEIGHTS
Legally, if I had a Creative Commons Share-Alike license on my data, and the LLM was trained on it and then served unlimited requests to others, without making the weights available…
…that would be almost exactly like if I had made my code available with Affero GPL license, someone would take my code but then incorporated it into a backend software hosting a social network or something, without making their own entire social network source code available. Technically this should be enforceable via a court order compelling the open sourcing to the public. (Alternatively, they’d have to pay damages in a class action lawsuit and stop using the tainted backend software or weights when serving all those people.)
TECHNICAL ANALYSIS
The key, as many here have missed, is authentication and authorization. You may have authorization to log in and view movies on Netflix. Not to rebroadcast them. Even the question of a VCR for personal use was debated in the past.
Distributing your scripts and software to process data is not the same as distributing arbitrary data the user agent found on the internet for which you don’t have a license.
If someone wrote an article, your reader transforms it based on your authenticated request, and your user would have an authorized subscription.
LEGAL ANALYSIS
Much of the content published on the Web isn’t secured with subscriptions and micropayments, which is why the whole thing becomes a legal battle as silly as “exceeding authorized access” which landed someone like Aaron Swartz in jail.
In other words, it is the question of “piracy”, which has acquired a new character only in that the AI is trained on your data and transforms it before it republishes it.
There was also a lawsuit aboot scraping LinkedIn, which was settled as follows: https://natlawreview.com/article/hiq-and-linkedin-reach-prop...
Legally, you can grant access to people subject to a certain license (eg Creative Commons Share Alike) and then any derived content must have its weights opened. Similar to, say, Affero GPL license for derivative software.
That is quite an important difference. IMHO.
"I don't want to live in a world where website owners can use DRM to force me to display their website in exactly the way that their designers envisioned it. I want to be able to write scripts to manipulate the page and present it in a way that's useful for me."
This is fundamentally different from a statement like, "I want to to able to access a third party server that accesses someone else's website, manipulates the pages and presents them in a way chosen by the third party that I find appealing."
These third party intermediaries, like Perplexity.ai, are not the same as a www user who retrieves a page from a website and reformats it to a suitable presentation, according to her own particular tastes, for her own consumption. The www user is not republishing or otherwise disseminating the reformatted page, and not for the purpose of generating web traffic, collecting data and profiting from the sale of advertising services. The www user may not even have their own website. Most www users do not have a website, let alone a search engine like Perplexity.ai.
There is no relationship between (a) the basic freedom of www users to choose their own www client, whether it is W3C linemode, Google Chrome or something the user wrote themselves, and (b) a third party intermediary that retrieves and redistributes/republishes other peoples' content for profit, without permission. Perplexity.ai is a server, a search engine, a website, not an http client.
The first concern is the most legitimate one: can I stop an LLM from training itself on my data? This should be possible and Perplexity should absolutely make it easy to block them from training.
The second concern, though, is can Perplexity do a live web query to my website and present data from my website in a format that the user asks for? Arguing that we should ban this moves into very dangerous territory.
Everything from ad blockers to reader mode to screen readers do exactly the same thing that Perplexity is doing here, with the only difference being that they tend to be exclusively local. The very nature of a "user agent" is to be an automated tool that manipulates content hosted on the internet according to the specifications given to the tool by the user. I have a hard time seeing an argument against Perplexity using this data in this way that wouldn't apply equally to countless tools that we already all use and which companies try with varying degrees of success to block.
I don't want to live in a world where website owners can use DRM to force me to display their website in exactly the way that their designers envisioned it. I want to be able to write scripts to manipulate the page and present it in a way that's useful for me. I don't currently use llms this way, but I'm uncomfortable with arguing that it's unethical for them to do that so long as they're citing the source.