Loading summary
A
Today on the podcast, I want to talk about a fascinating new AI company out of France called Linkup that is connecting LLMs to premium content. And they're doing this all legally. You can think of all of the drama around this as kind of being somewhat related to the lawsuits that the New York Times has had with OpenAI, but there's a lot more to it and this is a very interesting company. They've just raised a bunch of money. I want to get into all of that. Before I do, I wanted to mention if you have ever wanted to start a podcast, I would love to have you get my podcast studio course for Black Friday and Cyber Monday. I have a 50% off discount code. It'll be in the description. It's all caps altogether Black Friday, um, usually my course is 300 and it covers exactly how I started four top ranking podcasts and got over 4 million downloads on them. I go over all my strategies, everything I do, from content creation to the equipment I use to the marketing channels. There's a ton of really great things in the course curriculum there, and I would love to have you as a member. I think that podcasts are the number one way to raise your credibility as a creator, as a business owner, and for whatever project you're doing to get customers. So I'd love to have you as a member. I'm not doing this, you know, I'm not doing discounts very often. But for Black Friday Cyber Monday, it'll be 50% off, so you can get it over there. All right, let's get back into the podcast. So what's interesting is that you, you probably noticed from ChatGPT search and perplexity that these AI models, I think they do a fantastic job when they're, when they're working with the Internet, right when they're going and querying. There's a lot of questions that I ask that's especially about current events or things that are happening now where I just need like the newest information. Perplexity's done a great job with things, you know, around during the election. For example, they had a whole election portal which was updating who is winning seats, where, who is ahead in the polls, who is behind. And it's kind of like this live stuff and that's what you really want from any experience. Because Obviously when these AI models just had their 2021, you know, September 2021 cutoff window, it was quite limiting to anything new or that had happened in the last years. But even now, when you have closer cut off windows, anything that's happened in the last few months. Like, this is stuff we're interested in. So this French startup, Link up, they're building an API that essentially lets developers get access to web content, but also and especially premium web content from, you know, sources that they can include into their LLMs. So they're you. They're using REG, which is retrievement retrieval, augmented generation to do this. And I think this is absolutely fascinating. I've had a look over at their website at some of the things that Link up is doing. But you can imagine with all of the lawsuits that have happened and link ups making this very easy, five lines of code to, you know, get started with their product. But so they're making this very, very easy. But you can imagine with all the lawsuits that have happened, you know, New York Times suing OpenAI because they were scraping their content. There was, you know, a whole moment right after, I think GPT 3.5 may have launched or maybe GPT 4, where essentially if you asked ChatGPT to, you know, you could give it a link to a URL of premium content on New York Times and Chat GPT, I think, had purchased like a premium subscription. But, uh, if you just, if you give it the premium link and said, hey, summarize this article, it would summarize it for you. So obviously this was not great. The New York Times sued them about that because they're like, hey, look, you're giving away our paywall premium content for free.
B
But.
A
And so essentially, ChatGPT literally was off the Internet. Like they, they pulled the browse the Internet feature off for like six months while they fixed this, which was super annoying because it definitely kind of nerfed its capabilities. And you know, there's a lawsuit and all that kind of stuff that happens. But the problem is, even once they kind of came back on and we're like, okay, don't worry, we're not going to get onto your stuff. The issue is that, well, what if I want the New York Times? Like, how do I get that? Like, that's, you know, a great data set. That's very interesting. And so that's where Linkup comes in. They're going to essentially have APIs where you can get access to that payroll content. They're going to be paying for it. They have all the payments needed. So either the user, the company, or whoever is going to still be paying the company, like New York Times is still getting paid. What's really interesting though is beyond just an API to the New York Times, for example, they're actually creating a marketplace. So not just, you know, you can see like these big companies like New York Times that can do these lawsuits against OpenAI, but there's a lot of companies, you know, smaller publishers that can't go and fight these big AI models, right? Because either their choices, you know, say put a robot txt text on their website that says don't scrape us. So no search engine and also so, you know, no AI model but also no search engine can scrape them. You're not going to show up in any search results, you're not going to get a lot of traffic or you can just let yourself get scraped. Or you could try paywall. Like it's really, really tricky. So for smaller publishers that can't afford or afford to fund these big huge lawsuits, they can partner with some, a company like Link up who can add them to their marketplace. And now if I'm an AI company or even a developer that is using AI companies, right? So for example, myself, I'm not, you know, I don't have a foundational AI model, but I'm using all the foundational AI models on my platform AI box and they, you know, people, my users are, can go and pick what AI model they want to chat with and they can talk with them all and kind of use them all, all the, all the models in one chat. So super cool. But the problem is without these, you know, great data sources, they still don't get the best content. So by integrating with something like Link Up, I can now let my users, you know, as a third party developer, get access to all of this data that not only, you know, can they.
B
Not get access to over on ChatGPT, but now I'm adding kind of an extra, you know, an extra benefit and you can do a bunch of different packages. You can pay $400 and get a certain amount of tokens that essentially you can integrate into your tool. So for developers this is an amazing tool. Overall, I think this is a really awesome company there. It's interesting listening to kind of what their founders have to say about it. So the founding team said, quote, we're really targeting applications that are implementing AI in their own products. This was Missouri, one of their co founders and he said, so the typical use case is that I can create an AI application using a model for Mistral or OpenAI. I build my own pipeline, but I need to enrich this pipeline with external information, right? So all of this external, usually premium data is what they're kind of targeting on and targeting here. They said there's an example I like to use, which is one of our customers built an internal application for their salespeople. He said, on the one hand, they've listed all of the advantages of their own product and thanks us they get Reef or and thanks to us, they get fresh quality information on their prospectus and they put it into Mistral LLM. And Mistral LLM is going to generate a sort of sales pitch for sales reps, which they'll have in front of them when they make the calls with the customer sales. So really what it allows you to do is put in any information you want or get access to APIs of, you know, like the New York Times and all these other places. So I think this is really awesome. A very interesting company. There is competition in this field, which I will mention. The biggest one is Scale Post, which is a startup that's currently helping Perplexity to speed up their licensing deals with publishers. So there are some other players in this space, but I have a lot of confidence in Link up and I'm excited to see kind of where they go. They just raised 3 million euros for their seed round. It's about $3.2 million US and they got it from Alexo Capital, Monteer Ventures, Seed camp and about 100 business angels, which I think is kind of cool that they kind of did that. Almost like a crowdfund of business angels. There is about 10 people that are currently working for them and they're planning to hire another 10 people over the next year. All in all, a very interesting company. Very excited. I'll keep you up to date on everything happening with them. Thanks so much for tuning into the podcast. If you enjoyed it, if you could leave a review, it would mean the world to me. And also if you're interested in starting your own podcast again, there's a link in the description with our Cyber Monday code and Black Friday code. So it's just all caps Black Friday, but the codes in the description. If you want to get 50% off my podcasting course, which will seriously help you launch a successful podcast. So I'm excited to see what you start and if you do, make sure to message me on LinkedIn with the name of your podcast. Thanks so much and have a great rest of your day.
Joe Rogan Experience for AI
Episode: Linkup Raises $3.2M to Give LLMs Paywalled Articles
Release Date: December 8, 2024
In this episode of the "Joe Rogan Experience for AI," the host delves into the innovative strides of Linkup, a burgeoning AI startup from France. Linkup is making waves by legally connecting Large Language Models (LLMs) to premium, paywalled content—a solution poised to address significant challenges in the AI and publishing industries.
Linkup has emerged as a pivotal player aiming to bridge the gap between AI models and premium content sources. By developing an accessible API, Linkup enables developers to integrate high-quality, paywalled web content into their LLMs seamlessly. This approach not only enriches the data available to AI models but also ensures that content creators and publishers are fairly compensated.
The podcast highlights the contentious relationship between AI developers and content publishers, exemplified by the lawsuit between The New York Times and OpenAI. The host explains:
A (00:50): "You could give it a link to a URL of premium content on New York Times and ChatGPT... would summarize it for you. The New York Times sued them because they were giving away our paywall premium content for free."
This scenario underscores the legal and ethical dilemmas surrounding AI models' ability to access and utilize paid content without appropriate licensing or compensation.
Linkup addresses these challenges through its API, which leverages Retrieval-Augmented Generation (RAG). This technology allows AI models to access and incorporate premium content legally by ensuring that publishers are compensated for their content. The host elaborates:
A (02:15): "They're building an API that essentially lets developers get access to web content, but also and especially premium web content from sources that they can include into their LLMs."
This method not only provides AI models with up-to-date and high-quality data but also fosters a sustainable ecosystem where content creators can monetize their work.
One of Linkup's standout features is its creation of a marketplace that connects publishers of all sizes with AI developers. This platform democratizes access to premium content, enabling smaller publishers to participate without the intimidating prospect of legal battles against large AI companies. The host explains:
A (04:10): "For smaller publishers that can't afford or fund these big huge lawsuits, they can partner with a company like Linkup who can add them to their marketplace."
This marketplace model ensures that smaller content providers gain visibility and revenue streams, leveling the playing field in the digital content ecosystem.
Linkup's integration capabilities offer significant benefits to AI developers. By providing access to diverse and high-quality content, developers can enhance their applications' performance and reliability. The host shares a practical example:
A (04:55): "One of our customers built an internal application for their salespeople. Thanks to Linkup, they get fresh quality information on their prospects and integrate it into their LLM, generating effective sales pitches for their reps."
This integration not only improves the functionality of AI applications but also adds tangible value to end-users by providing relevant and timely information.
While Linkup is a frontrunner in this niche, the podcast acknowledges the presence of competitors like Scale Post. Scale Post is currently aiding Perplexity in accelerating their licensing deals with publishers. Despite the competition, confidence in Linkup remains high due to its robust approach and growing market presence.
A (05:20): "There is competition in this field, which I will mention. The biggest one is Scale Post... but I have a lot of confidence in Linkup and I'm excited to see where they go."
Linkup has successfully raised €3 million (approximately $3.2 million USD) in their seed round, securing investments from Alexo Capital, Monteer Ventures, Seedcamp, and around 100 business angels. This diverse investment base not only validates Linkup's business model but also positions them for substantial growth. The startup currently employs around 10 people with plans to double its workforce in the coming year, indicating strong expansion ambitions.
The founders emphasize their target market and the practical applications of their technology. One co-founder articulates:
A (06:05): "We're really targeting applications that are implementing AI in their own products. For example, creating an AI application using a model for Mistral or OpenAI, enriching it with external premium data."
These insights highlight Linkup's strategic focus on enhancing AI applications across various industries by providing essential data enrichment services.
The episode wraps up with a positive outlook on Linkup's trajectory. The host expresses enthusiasm about the company's potential to transform how AI models interact with premium content while ensuring fair compensation for content creators. Linkup is poised to play a crucial role in shaping the future of AI-driven applications and the digital content landscape.
A (07:10): "All in all, a very interesting company. Very excited. I'll keep you up to date on everything happening with them."
For listeners interested in the intersection of AI, technology, and business, this episode provides a comprehensive overview of Linkup's innovative approach to solving a pressing issue in the AI ecosystem. By facilitating legal access to premium content, Linkup not only enhances AI capabilities but also fosters a more equitable digital content environment.
Notable Quotes:
A (00:50):
"The New York Times sued them because they were giving away our paywall premium content for free."
A (02:15):
"They're building an API that essentially lets developers get access to web content, but also and especially premium web content from sources that they can include into their LLMs."
A (04:55):
"One of our customers built an internal application for their salespeople. Thanks to Linkup, they get fresh quality information on their prospects and integrate it into their LLM, generating effective sales pitches for their reps."
A (06:05):
"We're really targeting applications that are implementing AI in their own products. For example, creating an AI application using a model for Mistral or OpenAI, enriching it with external premium data."
A (07:10):
"All in all, a very interesting company. Very excited. I'll keep you up to date on everything happening with them."
This comprehensive summary encapsulates the key discussions, insights, and conclusions from the episode, providing a clear and engaging overview for those who haven't listened to the podcast.