<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Alien Intelligence on Medium]]></title>
        <description><![CDATA[Stories by Alien Intelligence on Medium]]></description>
        <link>https://medium.com/@alien.papers?source=rss-e5b0a7c9d40d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*4KbO9MOJTIEEiNXWkxb5QA.png</url>
            <title>Stories by Alien Intelligence on Medium</title>
            <link>https://medium.com/@alien.papers?source=rss-e5b0a7c9d40d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 05:28:29 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@alien.papers/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The value of your data is not token-worth]]></title>
            <link>https://papers.alien.club/the-value-of-your-data-is-not-token-worth-5d43c8ee8de0?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/5d43c8ee8de0</guid>
            <category><![CDATA[ai-rights-management]]></category>
            <category><![CDATA[data-valuation]]></category>
            <category><![CDATA[data-provenance]]></category>
            <category><![CDATA[tokenization-fallacy]]></category>
            <category><![CDATA[alien-intelligence]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Wed, 23 Apr 2025 15:32:40 GMT</pubDate>
            <atom:updated>2025-04-23T15:32:40.022Z</atom:updated>
            <content:encoded><![CDATA[<p><em>by Primavera De Filippi</em></p><h3>The fallacy of of token-based data valuation in AI</h3><p>In the context of generative AI, data contribution is reduced to a simplistic, mechanistic calculation based on the number of ‘tokens’ ingested by a model. The valuation is brutally simple: How many words? How many tokens? Nothing more.</p><p>Under this token-counting regime, all inputs are treated as functionally equivalent. The AI model doesn’t distinguish between scholarly depth and casual conversation, verified scientific research and unverified personal opinions, peer-reviewed content and crowd-sourced information, original intellectual work and derivative commentary. This means a meticulously researched PhD dissertation — representing years of rigorous academic work, original research, and deep intellectual contribution — is valued identically to a random collection of social media posts, internet comments, or viral tweets.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*lH_crzxS60rVSPW_.png" /><figcaption><a href="https://app.alias.studio/artist/MLIBTY/alias?aliasId=65807f2bffc9f7233dc3aa74">Image created with Impossible Papertoys by MLIBTY on Alias</a></figcaption></figure><p>This approach not only devalues intellectual labor but actively disincentivizes the creation or contribution of high-quality, nuanced content. Why invest years in rigorous research when a few thousand viral tweets can contribute the same “value” to an AI model?</p><p>In order to create more advanced and sophisticated LLMs, we must move away from consuming massive quantities of undifferentiated text — and instead start focusing on meaningful, high-quality contributions that genuinely advance knowledge and understanding. This requires AI model trainers to recognize that a single, ingeniously crafted dataset can be exponentially more valuable than massive volumes of generic information and that unique perspectives, creative insights, and specialized knowledge deserve proportional recognition.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/668/1*yVK2sFcKDCZF8DlzEGJxdg.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/668/1*o3IrQeNDeDsbMpzkcLqaJg.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/668/1*IdWggicrpvo7TvrHoISpOw.jpeg" /><figcaption><a href="https://app.alias.studio/artist/MLIBTY/alias?aliasId=65807f2bffc9f7233dc3aa74">Images created with Impossible Papertoys by MLIBTY on Alias</a></figcaption></figure><h3>Beyond tokenization: The true value of data</h3><p>The fundamental premise is straightforward yet profound: the value of your data should not be token-based. It’s not about how many units of information you provide, but the depth, originality, and transformative potential of that data.</p><p>While the law is lagging behind, technological solutions can and should create transparent mechanisms for attributing and rewarding these meaningful contributions. In particular, with AI-RM (AI Rights Management), the old paradigm of treating data as a quantitative commodity — simply counting tokens or data points — can give way to a more nuanced, qualitative approach, where data is evaluated through a multidimensional lens, based on a multiplicity of qualitative factors — e.g. the originality of insight, the depth of analysis, the verifiability of information, the expertise of the source, or the unique perspective contributed.</p><p>AI-RM represents a holistic approach to managing rights in the AI ecosystem. In addition to providing legal and technological tools to protect intellectual property, AI-RM also provides automated attribution systems, with transparent tracking of data provenance, as well as dynamic reward mechanisms that go beyond simple volume-based compensation.</p><h3>Microsoft’s attempts at data attribution</h3><p>Microsoft is pioneering a <a href="https://techcrunch.com/2025/03/21/microsoft-is-exploring-a-way-to-credit-contributors-to-ai-training-data/">groundbreaking research project </a>based on the concept of “training-time provenance” which could constitute an important breakthrough in the way we acknowledge human creativity behind artificial intelligence.</p><p>At the heart of Microsoft’s initiative is a revolutionary concept: tracking how specific pieces of data — whether photographs, written works, or artistic creations — directly influence the final output of an AI system. This goes far beyond simple token counting. As detailed in a <a href="https://www.theladders.com/job/research-intern-training-time-provenance-data-dignity-microsoft-mountain-view-ca_79951883">LinkedIn job listing for a research intern</a>, the goal is to develop a sophisticated method of estimating the precise contributions of various data sources.</p><p>Key insights from the research project include:</p><ul><li><strong>Transparent Data Attribution</strong>: The research aims to shed light on the typically opaque process of how neural networks amalgamate diverse inputs. This means being able to trace exactly how a particular artist’s style, a specific writer’s narrative technique, or a unique photographic composition might influence an AI-generated output.</li><li><strong>Inspiration from Data Dignity</strong>: The project draws heavily from the work of technologist Jaron Lanier, who advocates for “data dignity” — the principle that every digital artifact carries a human touch that deserves recognition. Imagine an AI generating an animated movie that draws inspiration from various artists — Microsoft wants to create a system that can precisely track and credit those inspirational sources.</li></ul><p>The research is not just about providing proper recognition — it’s also about creating a framework for equitable compensation, changing the way we view creative contributions to generative AI. Indeed, as this <a href="https://techcrunch.com/2025/03/21/microsoft-is-exploring-a-way-to-credit-contributors-to-ai-training-data/">article</a> notes, this approach is not just a technical upgrade — it’s a cultural shift that recognizes the human element behind every line of code and every stroke of digital art. By developing methods to track and potentially compensate data contributors, the company is laying the groundwork for a more just and transparent AI ecosystem, marking a promising step toward reconciling technological advancement with the fundamental rights of content creators.</p><p>This initiative comes at a crucial time, with numerous legal battles surrounding AI training data. Companies like The New York Times have filed <a href="https://www.nytimes.com/2023/12/27/business/media/new-york-times-open-ai-microsoft-lawsuit.html">lawsuits</a> arguing that their copyrighted content was used without permission. Microsoft’s approach offers a proactive solution — a transparent system that could potentially mitigate these legal challenges while creating a more equitable ecosystem for content creators.</p><p><strong>Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><p><strong>Follow Primavera de Filipi<br></strong><a href="https://x.com/yaoeo">X</a></p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><a href="https://aliasstudio.medium.com/"><strong><em>Medium Alias by Alien</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5d43c8ee8de0" width="1" height="1" alt=""><hr><p><a href="https://papers.alien.club/the-value-of-your-data-is-not-token-worth-5d43c8ee8de0">The value of your data is not token-worth</a> was originally published in <a href="https://papers.alien.club/">alien.papers</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[When AI Creates Art: The urgent need legal clarity in copyright law]]></title>
            <link>https://papers.alien.club/when-ai-creates-art-the-urgent-need-legal-clarity-in-copyright-law-d52fb725197f?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/d52fb725197f</guid>
            <category><![CDATA[ai-art]]></category>
            <category><![CDATA[ai-rights-management]]></category>
            <category><![CDATA[intellectual-property]]></category>
            <category><![CDATA[ai-and-creativity]]></category>
            <category><![CDATA[copyright-law]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Wed, 09 Apr 2025 07:41:21 GMT</pubDate>
            <atom:updated>2025-04-09T07:52:38.477Z</atom:updated>
            <content:encoded><![CDATA[<p><em>by Primavera de Filippi</em></p><p>In a landmark ruling, a U.S. federal appeals court recently decided that <a href="https://www.cnbc.com/2025/03/19/ai-art-cannot-be-copyrighted-appeals-court-rules.html">art created entirely by AI cannot be copyrighted</a>. The case, involving a painting called “A Recent Entrance to Paradise” created by Stephen Thaler’s “Creativity Machine,” highlights a critical challenge in copyright law: Who owns AI-generated content?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0KXNzHjk39zD2GOWCMXgUg.png" /><figcaption>A picture titled “A Recent Entrance to Paradise,” which computer scientist Stephen Thaler said he had created using his “Creativity Machine’’ AI platform.</figcaption></figure><h3>The controversy</h3><p>Computer scientist Stephen Thaler argued that the digital painting autonomously created by his AI system deserved copyright protection. However, the court unanimously rejected this claim, asserting that copyright law requires human authorship. The judges were clear: an AI cannot be an “author” in the legal sense.</p><p>The court acknowledged that AI is rapidly producing original work across industries and creative fields. Yet, this does not necessarily mean that these works are eligible for copyright protection. In fact, the current legal framework of copyright law struggles to address this new reality, with many open questions that remain unanswered. Is a work partially generated by AI copyrightable? What if a human provides significant creative direction?</p><h3>The need for legal clarity</h3><p>Stephen Thaler’s case is far more than a simple legal dispute — it’s a testament of the existential challenges facing creators in the age of artificial intelligence. By arguing that his “Creativity Machine” autonomously generated “A Recent Entrance to Paradise,” Thaler challenged fundamental assumptions about creativity, ownership, and economic value in a world where machines can already generate art that is indistinguishable from human-created works.</p><p>Consider the economic landscape emerging around AI-generated content:</p><ul><li>Visual artists who spent years developing distinctive styles now see those styles rapidly ingested and reproduced by AI models</li><li>Musicians find their entire catalogs used as training data without compensation</li><li>Writers discover their prose becoming fuel for language models that can generate derivative works at unprecedented speeds</li></ul><p>The court’s ruling — that human authorship remains a prerequisite for copyright — creates a precarious economic environment, suggesting that the value of creative work lies not in the output, but in the human hand guiding the process of creation.</p><p>Without a clear legal framework, artists may become reluctant to engage with AI technologies, creative professionals might view AI as an existential threat rather than a collaborative tool, and smaller creators could be systematically disadvantaged, unable to compete with well-resourced AI platforms.</p><p>The Thaler case is not just about whether an AI can be an “author” in the legal sense. It is about reimagining the entire ecosystem of creative production in the age of artificial intelligence, while establishing a precedent for how we value creative labor in an era of machine learning.</p><p><strong>Why we need AI Rights Management</strong></p><p>Paradoxically, the solution isn’t to resist AI but to create economic models that properly reward human creativity. The Thaler case illuminates a broader challenge: how do we create economic sustainability for creators in an AI-driven world? The current copyright framework, designed for a pre-digital era, simply cannot accommodate the complexity of AI-generated content.</p><p>AI Rights Management (AI-RM) transforms this challenge into an opportunity. Instead of seeing AI as a threat, we could view it as a new collaborative ecosystem where human creativity is systematically valued and rewarded.</p><p>Specifically, with AI-RM, we can create new economic frameworks where artists can license their style or previous works for AI training, and receive micro-payments when their work contributes to AI-generated content.</p><p>AI-RM to enable us to:</p><ul><li><strong>Clarify Ownership</strong>: Establishing clear attribution mechanisms and guidelines for who owns AI-generated content</li><li><strong>Protect Creators</strong>: Creating transparent compensation models to ensure that artists and original content creators are fairly compensated</li><li><strong>Enable Innovation: </strong>Creating a system that allows technological progress without undermining creative rights, by recognizing the nuanced contributions of both human creators and AI systems</li></ul><p>Ultimately, by establishing clear economic frameworks, we don’t just protect existing creators — we incentivize future innovation. Creators will be more likely to experiment with AI tools if they understand how their intellectual contributions will be recognized and compensated.</p><p>The alternative — a legal and economic Wild West where creators feel exploited — risks stifling the very creativity that makes technological innovation meaningful.</p><h3>What Next?</h3><p>The current ruling creates what Thaler’s lawyer called “a huge shadow on the creative community” by leaving critical questions unanswered. Thaler thus plans to appeal, potentially taking the case to the Supreme Court. This suggests the legal community recognizes the profound implications of this debate.</p><p>As AI continues to transform creative industries, one thing becomes clear: we need a new paradigm for understanding ownership, creativity, and managing rights in the age of artificial intelligence. As copyright law cannot cope with the pace of technological innovation, there is an urgent need for a more nuanced approach that:</p><ul><li>Recognizes the collaborative nature of AI creation</li><li>Provides clear guidelines for copyright</li><li>Ensures fair compensation for human creators</li><li>Enables continued technological innovation</li></ul><p>AI Rights Management represents our best hope for navigating these complex waters. The challenge is not about choosing between human creativity and AI capabilities, but about creating a framework where both can thrive.</p><p><strong>Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><p><strong>Follow Primavera de Filipi<br></strong><a href="https://x.com/yaoeo">X</a></p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><a href="https://aliasstudio.medium.com/"><strong><em>Medium Alias by Alien</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d52fb725197f" width="1" height="1" alt=""><hr><p><a href="https://papers.alien.club/when-ai-creates-art-the-urgent-need-legal-clarity-in-copyright-law-d52fb725197f">When AI Creates Art: The urgent need legal clarity in copyright law</a> was originally published in <a href="https://papers.alien.club/">alien.papers</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Explain me like I’m five: AI Rights Management]]></title>
            <link>https://medium.com/@alien.papers/explain-me-like-im-five-ai-rights-management-a871d54bc0ee?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/a871d54bc0ee</guid>
            <category><![CDATA[intellectual-property]]></category>
            <category><![CDATA[ai-rights-management]]></category>
            <category><![CDATA[explain-like-im-5]]></category>
            <category><![CDATA[rmai]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Tue, 01 Apr 2025 07:54:33 GMT</pubDate>
            <atom:updated>2025-04-01T07:54:33.555Z</atom:updated>
            <content:encoded><![CDATA[<p><em>by Primavera De Filippi</em></p><p>Imagine you’re in an art class where everyone brings their own drawings, photos, and paintings to share and learn from. Some kids also bring pictures of themselves and their families. One day, your teacher brings in a special robot artist that can look at all these pictures and create new artwork. That sounds amazing, right?</p><p>But wait — just like how you wouldn’t want someone copying your drawing without asking, or using a photo of you without permission, we need rules for how this robot artist can use everyone’s work and pictures. This is where AI Rights Management (or AI-RM for short) comes in!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Hi3Odx1hK6JKGxzH.png" /><figcaption>Image created with the help of <a href="https://app.alias.studio/artist/Dellas/alias?aliasId=6690eb880cdfea6cce7f57a3">Oil Dreams by Dellas on Alias Studio</a></figcaption></figure><h3>What is AI Rights Management?</h3><p>Think of AI-RM as your art teacher who makes sure everyone’s artwork and photos are used fairly. Just like how your teacher keeps track of whose drawing is whose and asks permission before displaying someone’s art in the school hallway, AI-RM helps make sure the robot artist follows similar rules.</p><h3>Why do we need it?</h3><p>Remember when someone once copied your artwork without asking, and you felt upset? Well, the robot artist can sometimes accidentally do something similar because it learned by looking at lots of different artworks. AI-RM helps by:</p><ol><li>Making sure the robot artist has permission to look at and learn from certain pictures</li><li>Keeping track of whose art styles and photos were used for teaching the robot</li><li>Making sure the robot’s new artwork doesn’t copy too much from any one person’s art</li></ol><h3>How is it different from old rules?</h3><p>The old way of protecting art (called DRM) was like putting your drawing in a locked box where only you could see it — even if your best friend just wanted to learn from your cool coloring technique. AI-RM is different. It’s more like having a special art journal that shows:</p><ul><li>Which artists gave permission for the robot to learn from their work</li><li>What kinds of pictures the robot is allowed to make</li><li>How we can share and use the robot’s artwork fairly</li></ul><h3>What makes AI-RM special?</h3><p>AI-RM helps in three important ways:</p><h3>1. It’s like a chain of title</h3><p>Just like how when you buy a house, you want to make sure the seller actually owns it and has the right to sell it, AI-RM keeps track of who has given permission for their creative work to be used. It creates a clear record showing who owned what and who gave permission for what — just like the paper trail that proves you really own your home.</p><h3>2. It protects everyone’s privacy</h3><p>Think of it like your personal medical records or financial information — you wouldn’t want just anyone accessing or sharing those. Similarly, AI-RM ensures that AI systems don’t use anyone’s personal data, private photos, or sensitive information without proper consent and safeguards in place.</p><h3>3. It helps create original art</h3><p>AI-RM is like having an art teacher who helps you develop your own style instead of just copying others. It guides the robot artist to create new things while respecting everyone else’s work.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*brLlzHVgFzTF3unH0uQQQA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*7rvSpYNf4ABlYNy0bsXiSg.png" /><figcaption>Image created with the help of <a href="https://app.alias.studio/artist/Dellas/alias?aliasId=6690eb880cdfea6cce7f57a3">Oil Dreams by Dellas on Alias Studio</a></figcaption></figure><h3>What is AI Rights Management?</h3><p>Think of AI-RM as your art teacher who makes sure everyone’s artwork and photos are used fairly. Just like how your teacher keeps track of whose drawing is whose and asks permission before displaying someone’s art in the school hallway, AI-RM helps make sure the robot artist follows similar rules.</p><h3>Why do we need it?</h3><p>Remember when someone once copied your artwork without asking, and you felt upset? Well, the robot artist can sometimes accidentally do something similar because it learned by looking at lots of different artworks. AI-RM helps by:</p><ol><li>Making sure the robot artist has permission to look at and learn from certain pictures</li><li>Keeping track of whose art styles and photos were used for teaching the robot</li><li>Making sure the robot’s new artwork doesn’t copy too much from any one person’s art</li></ol><h3>How is it different from old rules?</h3><p>The old way of protecting art (called DRM) was like putting your drawing in a locked box where only you could see it — even if your best friend just wanted to learn from your cool coloring technique. AI-RM is different. It’s more like having a special art journal that shows:</p><ul><li>Which artists gave permission for the robot to learn from their work</li><li>What kinds of pictures the robot is allowed to make</li><li>How we can share and use the robot’s artwork fairly</li></ul><h3>What makes AI-RM special?</h3><p>AI-RM helps in three important ways:</p><h3>1. It’s like a chain of title</h3><p>Just like how when you buy a house, you want to make sure the seller actually owns it and has the right to sell it, AI-RM keeps track of who has given permission for their creative work to be used. It creates a clear record showing who owned what and who gave permission for what — just like the paper trail that proves you really own your home.</p><h3>2. It protects everyone’s privacy</h3><p>Think of it like your personal medical records or financial information — you wouldn’t want just anyone accessing or sharing those. Similarly, AI-RM ensures that AI systems don’t use anyone’s personal data, private photos, or sensitive information without proper consent and safeguards in place.</p><h3>3. It helps create original art</h3><p>AI-RM is like having an art teacher who helps you develop your own style instead of just copying others. It guides the robot artist to create new things while respecting everyone else’s work.</p><h3>Why should we care?</h3><p>As more people start using robot artists to create things — like stories, pictures, and music — we need to make sure everyone’s original work and personal information is protected. AI-RM helps by:</p><ul><li>Making sure artists get credit when their style inspires new artwork</li><li>Keeping private pictures and information safe</li><li>Making sure the robot artist creates things in a way that’s fair to everyone</li></ul><h3>The future is bright!</h3><p>Just like how art class is more fun when everyone respects each other’s work and shares fairly, AI-RM makes it easier and safer for people to use AI to create new things. It helps everyone feel confident about sharing their art and ideas with robot artists, knowing there are good rules to protect their work and privacy.</p><blockquote>Remember: AI-RM isn’t about stopping creativity — it’s about making sure everyone’s rights are respected while we create amazing new things together!</blockquote><p><strong>Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><p><strong>Follow Primavera de Filipi<br></strong><a href="https://x.com/yaoeo">X</a></p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><a href="https://aliasstudio.medium.com/"><strong><em>Medium Alias by Alien</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a871d54bc0ee" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Studio Ghibli: When AI Meets Artistic Legacy]]></title>
            <link>https://papers.alien.club/studio-ghibli-when-ai-meets-artistic-legacy-fc65d004ad65?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/fc65d004ad65</guid>
            <category><![CDATA[studio-ghibli]]></category>
            <category><![CDATA[ai-art]]></category>
            <category><![CDATA[copyright]]></category>
            <category><![CDATA[alias-by-alien]]></category>
            <category><![CDATA[creative-rights]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Mon, 31 Mar 2025 10:26:30 GMT</pubDate>
            <atom:updated>2025-03-31T10:26:30.125Z</atom:updated>
            <content:encoded><![CDATA[<h3>The AI Controversy</h3><p>In just 24 hours, OpenAI’s latest image generation tool triggered a social media storm by effortlessly recreating the iconic Studio Ghibli aesthetic. From Elon Musk to “The Lord of the Rings” characters, users flooded platforms with AI-generated images mimicking the studio’s distinctive style — a style cultivated over decades by legendary animators like Hayao Miyazaki.</p><p>Studio Ghibli’s aesthetic is more than just an art style — it’s a carefully crafted visual language characterized by ethereal landscapes that blend reality and fantasy, with watercolor-like textures, intricate hand-drawn details, and emotional depth in character expressions.</p><p>OpenAI claims that their model prevents the replication of “individual living artists” styles. Yet, they permit the recreation of “broader studio styles”. This creates a murky legal and ethical boundary, especially for studios like Ghibli where individual artists (like Miyazaki) are intrinsically linked to the studio’s visual identity.</p><p>In a recent <a href="https://techcrunch.com/2025/03/26/openais-viral-studio-ghibli-moment-highlights-ai-copyright-concerns/">TechCrunch article</a>, intellectual property lawyer Evan Brown described the complex legal landscape of copyright law in the age of AI. Because style itself is not explicitly protected by copyright, OpenAI cannot be held liable of copyright infringement for generating images in the style of Ghibli. However, using copyrighted images in the training data is more contentious. On the one hand, it is unclear whether training an AI model with copyrighted content might fall under fair use exemptions. On the other hand, if the output of the model were to be substantially similar to some of the input images in the training dataset, there may be a legitimate claim for copyright infringement.</p><blockquote>Because style itself is not explicitly protected by copyright, OpenAI cannot be held liable of copyright infringement for generating images in the style of Ghibli. However, using copyrighted images in the training data is more contentious.</blockquote><p>In order to evaluate the potential liability of Open AI, it is thus necessary to assess whether OpenAI did train its model on millions of Ghibli film frames, and the extent to which such training on these copyrighted works constitute fair use — which also involve assessing whether the relevant artists have been fairly compensated for their distinctive efforts.</p><p>Indeed, the ability of OpenAI’s model to rapidly replicate Studio Ghibli’s aesthetic raises significant economic concerns, which is not limited to the devaluation of original artworks when AI-generated alternatives become widely available, but also extends to the potential loss of licensing revenue for studios as their distinctive styles become freely reproducible.<strong> The result might be a reduction of incentives for investment in developing unique artistic approaches, which might significant affect smaller studios and independent artists who rely on distinctive visual styles.</strong> As a result, while technology companies significantly profit from these AI models, the original creators whose work trained these systems may see diminishing returns on their artistic investments. <strong>This economic imbalance threatens the sustainability of traditional animation studios and may reduce stylistic diversity in the industry over time.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YjjmS3p_WiOGXX6RIqPlYQ.png" /><figcaption><a href="https://app.alias.studio/artist/Layla/alias?aliasId=67b44ddb198dd9ecfbd80675">Reference image for rainbow-diagrams model by Layla available on Alias.studio</a></figcaption></figure><h3>The Human Element</h3><p>But while AI can generate Ghibli-like images, it cannot replicate the emotional storytelling, the nuanced character development, the philosophical depth of Miyazaki’s work and the cultural context that informs his work. Because Miyazaki’s works aren’t just visually distinctive — they’re deeply embedded in Japanese cultural traditions, environmental philosophy, and post-war societal reflection. The aesthetic choices in films like “Princess Mononoke” or “Spirited Away” emerge from specific cultural touchpoints including Shinto animism, Japanese folklore, and the country’s complex relationship with industrialization and nature.</p><p>These cultural elements inform seemingly simple visual decisions: the reverence shown to elderly characters reflects Japanese respect for age and wisdom; the detailed portrayal of food preparation celebrates daily rituals; the architectural designs blend traditional Japanese structures with European influences that entered Japan during the Meiji period. Even the way characters move through space — the thoughtful pauses, the attention to mundane activities — embodies concepts like “ma” (negative space) and “mono no aware” (the pathos of things) that are fundamental to Japanese aesthetics.</p><p>Besides, Ghibli’s art is inseparable from Japan’s post-war identity and Miyazaki’s experiences growing up during this transformative period. His father’s work in the aviation industry during wartime directly influenced the complex portrayal of flight and technology throughout his films — at once celebratory and cautionary. The environmental themes reflect both ancient Shinto beliefs about nature spirits and modern Japanese reckoning with rapid industrialization.</p><p>AI systems, regardless of their technical sophistication, lack the lived experience that generates these cultural resonances. They can mimic the visual patterns but cannot authentically participate in the cultural dialogue that produced these works. Without understanding the historical trauma, cultural celebrations, and philosophical traditions that Miyazaki draws upon, AI-generated “Ghibli-style” images remain surface-level imitations divorced from the profound cultural context that gives the original work its emotional weight and significance.</p><p>As AI becomes more sophisticated, we must reimagine the relationship between human creativity and machine learning. The goal is not to restrict AI, but to ensure it becomes a tool that respects and elevates human artistic expression. This requires intentional design choices that prioritize human agency and creative contribution. Rather than positioning AI as a replacement for human artists or as a shortcut to mimicking established styles, we should envision AI systems that function as collaborative partners — tools that expand creative possibilities while preserving the essential human elements of art.</p><p>This requires developing AI frameworks that explicitly attribute and compensate artists whose styles influence the technology. For instance, with <a href="https://alias.studio/">Alias.studio,</a> we have created a systems that require meaningful human input and direction rather than merely generating outputs from simple prompts. Alias.studio is a co-creative interface where artists can use AI to explore variations of their own styles, overcome technical barriers, or generate novel starting points that they then refine and imbue with personal meaning.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8SQ5dho7iQGL8vKoNGwRaw.png" /><figcaption><a href="http://alias.studio/">Alias.Studio</a></figcaption></figure><h3>AI Rights Management: Navigating the Copyright Minefield of Creative AI</h3><p>Despite all this, the viral sensation of AI-generated Studio Ghibli-style images has brought a critical question into the spotlight: How do we protect creative rights in the age of generative AI?</p><p>The challenge of balancing artistic integrity with technological innovation requires thoughtful solutions that protect creators while enabling progress. What emerges from this tension is the need for a comprehensive AI Rights Management framework that can address the complex intersection of art, technology, and ownership.</p><p>First, it must protect original creators by ensuring artists receive fair compensation when their styles contribute to AI training or inspire AI-generated content. This protection acknowledges that artistic styles represent significant intellectual investments deserving recognition.</p><p>Second, effective rights management must enable continued innovation by providing clear guidelines that remove legal uncertainty. With established boundaries, AI developers can confidently push creative boundaries without fear of unexpected liability.</p><p>Third, this framework should create new revenue streams through licensing models where content creators can monetize their stylistic signatures — transforming what is currently an extraction of value into a mutually beneficial exchange.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zx5N06hEisj_dMlQ6DuH2Q.png" /><figcaption><a href="https://app.alias.studio/artist/CROSSLUCID">Infinite Assembler by CROSSLUCID on Alias.studio</a></figcaption></figure><h3>The Alias.studio Solution</h3><h4>1. Attribution and Certification</h4><p>Attribution on Alias.studio represents more than mere acknowledgment — it maintains the crucial connection between creators and their work. Artists and creators build their own generative AI models, and can certify it to ensure that the outputs of the model are consistent and coherent with their own artistic style. This system makes it possible to preserve artistic lineage while educating users about the human creativity that underlies even AI-generated works.</p><h4>2. Licensing and Royalty system</h4><p>The current opacity around how AI systems are trained creates both legal uncertainty and ethical concerns. Alias.studio implements transparent licensing requiring users to secure acquire proper authorization before using distinctive artistic styles. This approach creates clear revenue streams for artists and studios, acknowledging their creative contributions while providing users with more legal certainty. As such, the Alias.studio platform simplify compliance while ensuring fair compensation reaches creators, similar to how music licensing works across streaming platforms.</p><h3>The Path Forward: Balancing Innovation and Creative Rights</h3><p>The rapid advancement of AI image generation has created a technological frontier that often resembles the Wild West — a place of tremendous opportunity but limited governance. As we’ve seen with the Studio Ghibli case, this unregulated space can lead to tensions between technological progress and artistic legacy. Yet the path forward isn’t about restricting AI’s creative potential but rather channeling it through frameworks that respect the foundation upon which it builds.</p><h3>Creating a Sustainable Ecosystem</h3><p>Effective AI Rights Management isn’t about stopping innovation; it’s about creating a fair ecosystem where technology and creativity can coexist and mutually benefit. This ecosystem requires a collaborative approach that simultaneously:</p><ul><li>Respects the intellectual property of creators by acknowledging the years of dedication behind distinctive artistic styles and ensuring appropriate compensation when those styles influence AI development</li><li>Encourages technological innovation by providing clear guidelines that remove legal uncertainty and allow AI developers to confidently explore new creative territories</li><li>Creates new economic opportunities for both artists and AI developers through licensing structures, attribution systems, and new forms of creative collaboration</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y_bA1KEAmDbjCByvWcr3Hg.png" /><figcaption><a href="https://app.alias.studio/artist/Gregorio%20Undurraga/alias?aliasId=656ccbdd015ba0970bc13e7e">Undurraga Oil Paintings v0.70 by Gregorio Undurraga on Alias.studio</a></figcaption></figure><h3>From Conflict to Collaboration</h3><p>The tension between traditional artistic creation and AI-generated content represents not an insurmountable conflict but rather a transition phase. Throughout history, new technologies have disrupted creative industries — from photography’s impact on portrait painting to digital tools transforming the music industry. In each case, initial resistance eventually gave way to integration, with the technology becoming another tool in the artist’s arsenal rather than a replacement.</p><p>The same evolution is possible with AI image generation. By developing thoughtful frameworks now, we can accelerate this transition from perceived threat to collaborative tool. Artists might leverage AI to explore variations of their own styles, overcome technical limitations, or generate starting points that they then refine with their unique human perspective and intention.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*glmYradkbAKKuroWFGGg3g.png" /><figcaption><a href="https://app.alias.studio/artist/jbardakos/alias?aliasId=6576f639a6df6805ed6a835d">FORMAFORMS by jbardakos on Alias.studio</a></figcaption></figure><h3>The Imperative for Action</h3><p>The AI revolution is already here — and with it comes the imperative to develop robust, flexible frameworks that protect creative rights while unleashing unprecedented creative potential. These frameworks must be adaptable enough to evolve alongside the technology they govern, recognizing that today’s solutions may need refinement as AI capabilities continue to advance.</p><p>Industry leaders, policymakers, artists, and technologists all have crucial roles to play in this development. By bringing diverse perspectives to the table, we can create systems that balance competing interests and find solutions that serve the broader creative ecosystem rather than any single stakeholder group.</p><p>It’s time to transform the Wild West of AI creativity into a structured, equitable landscape — one where the value of human creativity is recognized and preserved even as we embrace the extraordinary possibilities that AI brings to artistic expression. In this transformed landscape, we won’t just be protecting the past; we’ll be laying the groundwork for a future where human and artificial intelligence combine to create art forms we have yet to imagine.</p><p><strong>Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><p><strong>Follow Primavera de Filipi<br></strong><a href="https://x.com/yaoeo">X</a></p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><a href="https://aliasstudio.medium.com/"><strong><em>Medium Alias by Alien</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fc65d004ad65" width="1" height="1" alt=""><hr><p><a href="https://papers.alien.club/studio-ghibli-when-ai-meets-artistic-legacy-fc65d004ad65">Studio Ghibli: When AI Meets Artistic Legacy</a> was originally published in <a href="https://papers.alien.club/">alien.papers</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Explain me like I’m five: LoRA (Low Rank Adaption)]]></title>
            <link>https://papers.alien.club/explain-me-like-im-five-lora-low-rank-adaption-a7ce10c6d7c2?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/a7ce10c6d7c2</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[deep-learning-technology]]></category>
            <category><![CDATA[lora]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Fri, 21 Mar 2025 16:59:30 GMT</pubDate>
            <atom:updated>2025-03-31T11:25:29.801Z</atom:updated>
            <content:encoded><![CDATA[<p><em>by Primavera De Filippi</em></p><p>Imagine you’re a master chef who has spent years perfecting your recipes in a massive industrial kitchen. This kitchen is like a large language model or AI image generator — it’s powerful, expensive, and took a lot of time to set up perfectly. This is your “base model.”</p><p>Now, what if you want to adapt your kitchen to make a specific type of cuisine — let’s say Thai food? You could:</p><ol><li>Build an entirely new kitchen (training a new model from scratch)</li><li>Completely reorganize your existing kitchen (fine-tuning the entire model)</li><li>Add a small specialized cooking station just for Thai cuisine (using a LoRA)</li></ol><p>This is where Low-rank Adaptation (LoRA) comes in — it’s like adding that small specialised cooking station instead of rebuilding the entire kitchen.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yBR66tfi7dpa87NZrdMJAw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GUO36-gMeWywti4uCORsnA.png" /></figure><h3>Why is LoRA so clever?</h3><p>Think about what you really need to adapt your kitchen for Thai cooking:</p><ul><li>A few specific spices and ingredients</li><li>Some specialized cooking tools</li><li>New recipes and techniques</li></ul><p>You don’t need to:</p><ul><li>Change how the ovens work</li><li>Replace all your pots and pans</li><li>Rebuild the entire ventilation system</li></ul><p>Similarly, LoRA works by making small, strategic changes to an AI model instead of modifying everything. It’s like creating a compact “Thai cuisine upgrade kit” that can be easily added to or removed from your main kitchen.</p><h3>The Benefits of Using LoRAs</h3><h3>1. Efficiency</h3><p>Instead of needing another massive kitchen (which would require lots of space, money, and energy), you just need a small specialized station. LoRAs are tiny compared to full models — often less than 1% of the original model’s size.</p><h3>2. Flexibility</h3><p>Just like you could have different specialized stations for Thai, Italian, or Mexican cuisine that you can swap in and out, you can have multiple LoRAs for different specialties and switch between them easily.</p><h3>3. Preservation</h3><p>Your main kitchen (base model) stays intact. If the Thai cooking station isn’t working out, you can simply remove it without affecting your ability to cook other cuisines.</p><h3>How Does it Actually Work?</h3><p>Let’s stick with our kitchen metaphor:</p><ol><li><strong>Analysis</strong>: First, LoRA looks at how your kitchen operates when cooking Thai food. It’s like having an expert chef watch and identify exactly what needs to change to make authentic Thai cuisine.</li><li><strong>Efficient Changes</strong>: Instead of writing down every single detail, the chef creates a simplified set of instructions that capture the essential differences. In LoRA terms, this is called “low-rank decomposition” — it’s finding the smallest number of changes that will give you the biggest impact.</li><li><strong>Adaptable Instructions</strong>: These changes are written in a way that works with your existing kitchen setup. Just like how a Thai cooking station needs to connect to your existing power and water supplies, a LoRA needs to integrate smoothly with the base model.</li></ol><h3>Real-World Applications</h3><p>LoRAs are incredibly useful in many scenarios:</p><ol><li><strong>Personalization:</strong> Just like how you might adapt your kitchen for specific dietary requirements, LoRAs can customize AI models for particular users or use cases.</li><li><strong>Rapid Adaptation:</strong> Need to quickly switch between different cooking styles for different events? LoRAs let you rapidly adapt AI models for different tasks without lengthy retraining.</li><li><strong>Resource Efficiency:</strong> Running a full industrial kitchen is expensive. LoRAs let you achieve specialized results without the full cost of maintaining multiple complete kitchens.</li></ol><h3>The Future of LoRAs</h3><p>As AI continues to evolve, LoRAs are becoming increasingly important. They’re like having a set of specialized cooking stations that can be easily shared, combined, and modified. This makes advanced AI more accessible and customizable for everyone — not just those who can afford to build and maintain massive models.</p><p><strong>Remember</strong>: Just as a small, well-designed cooking station can transform a general kitchen into a specialized one, LoRAs can efficiently adapt powerful AI models for specific purposes without losing their original capabilities.</p><blockquote>The next time you hear about LoRAs, think of them as those clever, compact cooking stations that can transform a general-purpose kitchen into something specialized — all without having to rebuild the entire kitchen from scratch!</blockquote><p><strong>Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><p><strong>Follow Primavera de Filipi<br></strong><a href="https://x.com/yaoeo">X</a></p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><a href="https://aliasstudio.medium.com/"><strong><em>Medium Alias by Alien</em></strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a7ce10c6d7c2" width="1" height="1" alt=""><hr><p><a href="https://papers.alien.club/explain-me-like-im-five-lora-low-rank-adaption-a7ce10c6d7c2">Explain me like I’m five: LoRA (Low Rank Adaption)</a> was originally published in <a href="https://papers.alien.club/">alien.papers</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is the equivalent of streaming for generative AI?]]></title>
            <link>https://papers.alien.club/what-is-the-equivalent-of-streaming-for-generative-ai-45a4f5ed94b4?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/45a4f5ed94b4</guid>
            <category><![CDATA[online-media-consumption]]></category>
            <category><![CDATA[streaming]]></category>
            <category><![CDATA[hadopi-law]]></category>
            <category><![CDATA[copyright-law]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Tue, 04 Mar 2025 16:46:45 GMT</pubDate>
            <atom:updated>2025-03-04T16:46:45.991Z</atom:updated>
            <content:encoded><![CDATA[<p><em>by Primavera de Filippi</em></p><p>In the 1990s, the Internet was quite different from what we know today. It was the era when the web was just getting established, and we were gradually discovering the wonders of this new communication tool.</p><p>It was also the time when, suddenly, we discovered that we could digitize content (such as CDs or DVDs) and share it on the Internet with friends, but also obviously with strangers. And then, we discovered that all this content could not only be reproduced and transferred to the other side of the world in seconds but could also be modified and remixed in just a few clicks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-Z_swo_VPDEH0OicW3Qz6Q.png" /><figcaption>Image created using <a href="https://app.alias.studio/artist/XkMare">Mutants by XkMare on Alias</a></figcaption></figure><p>This was excellent news for those who wanted to maximize the visibility of this content and encourage greater consumption and production of cultural content, but it was bad news for copyright holders, who wanted to prevent unauthorized reproduction and distribution of their content. With the arrival of peer-to-peer file-sharing networks (such as BitTorrent, Gnutella, or Kazaa, for those who remember), the Internet began to be perceived, particularly by rights holders, as a threat to their business model.</p><p>With digital file piracy becoming accessible to everyone, no one would buy CDs or DVDs anymore, and the cultural industry would then collapse, as artists and content publishers would no longer be able to collect enough revenue to compensate for their efforts.</p><p>The first reaction was to strengthen copyright law by criminalizing the unauthorized sharing and downloading of protected digital content. In France, the <a href="https://en.wikipedia.org/wiki/HADOPI_law">HADOPI law</a> introduced a “graduated response” system going as far as suspending internet connection for those who would not respect copyright law. Although this may have discouraged few people, these legal reforms did not succeed in containing the negative effects of online piracy on creative industries’ revenues.</p><p>The private sector then took up the issue, developing technological solutions aimed at restricting the possibilities of reproduction and redistribution of digital content. This is the case with all anti-copy systems to prevent the digitization of CDs or DVDs. However, these technological solutions extended well beyond the scope of copyright, to encroach on the access and consumption of digital content itself: When you bought an MP3 on iTunes, you could only listen to it on Apple devices, and no more than 5 devices at once. And if you bought a DVD in the United States, you might not be able to play it on your computer in France. These usage restrictions are such that content “protected” by these technological solutions is of inferior quality to pirated content, which is freely accessible on the Internet. And this naturally contributes to reinforcing the advantages of online piracy.</p><p>The solution to piracy was not a legal reform nor a technological solution, but a commercial innovation. Most people who downloaded digital content in violation of copyright did not do so because they didn’t want to pay for the content, but because it allowed them to access a practically unlimited collection of content, accessible in a few clicks, and without any usage restrictions. This was more interesting than most commercial offers that existed at the time.</p><p>Eventually, a new commercial offer arrived with “streaming” — with platforms such as Deezer or Spotify providing internet users with an almost unlimited amount of content, accessible instantly, without having to download files to their devices. The film industry followed suit, with platforms like Netflix, which make available a large collection of movies or series, accessible at any time.</p><p>And suddenly, the Internet, which appeared as a threat to creative industries, translated into a new commercial opportunity, allowing passive income generation for rights holders. So, unfortunately, many artists are exploited by streaming platforms that hold a monopoly position, but the emergence of these platforms has nevertheless demonstrated that it is possible to monetize digital content, even that which is freely accessible on peer-to-peer networks, as long as you provide a superior quality service for which people are willing to pay.</p><p>And then, as usual, history repeats itself. Today, with Artificial Intelligence, we are experiencing a new period of disruption, with a new technology that — like the Internet — presents both opportunities and threats to creative industries. On one hand, Artificial Intelligence can generate thousands of personalized content pieces in seconds, and almost automatically. With artificial creativity increasingly rivaling human creativity. On the other hand, this automation of creativity questions the very foundations of copyright law, since the law only recognizes humans, not machines, as having the capacity to be authors. Any work generated by AI will therefore not be protected by copyright and will automatically belong to the public domain.</p><p>However, the real problem raised by rights holders is that AI threatens their business models, both through its efficiency (since AI can generate content much faster than a human) and its economy (since AI work costs much less than human work). Yet these artificial intelligences have been trained on huge databases, fed by works produced by artists worldwide. All cultural content accessible on the Internet has been used as training data to create increasingly powerful generative AI models, capable of producing an infinity of content, inspired by and derived from training data, which is often protected by copyright. But if art feeds the machines, who feeds the artists?</p><p>Last month, Meta admitted to downloading more than 100 terabytes of pirated books from BitTorrent (equivalent to half a million digital books) to train its artificial intelligence models. In the HADOPI era, this would have been considered a criminal activity punishable by 3 years in prison.</p><p>But the fundamental question remains whether training an AI on copyright-protected content actually constitutes copyright infringement? While some claim that any unauthorized processing of protected data should be considered “theft” of data, especially when this processing is done for commercial purposes, others claim that using data as a learning base for AI models only stores this data in their knowledge system, so they can later draw inspiration from it to create new works — Which should not, in theory, be protected by copyright.</p><p>Yet, while the legality of the training phase remains uncertain, the legality of the generation phase is much clearer: if the content generated by an AI model is too similar to the content the model has ingested, there will probably be an infringement. This poses a legal risk both to model producers and to users of these models, who risk generating illicit images unknowingly. Once again, the solution to this problem cannot be limited to legal reform, nor to an exclusively technological device, we need to seek the commercial innovation that will make the problem obsolete. In other words, we need to identify what is the equivalent of “streaming” in the AI era.</p><p>As a researcher at CNRS and Harvard University, I have dedicated my academic career to researching how copyright can coexist with digital technology, and how this medium can bring opportunities to rights holders, instead of threatening them.</p><blockquote>Today, I can’t help but try to find a solution regarding copyright and AI. And here is my conclusion: Generative AI systems suffer from two main problems: hallucinations and homogenization.</blockquote><p>Hallucinations are such that whatever we ask an AI, it will generate it for us, even if it doesn’t know what it’s doing. This is great in an artistic context, where hallucination translates into imagination; but it is problematic in any context where the veracity of information is important, particularly in legal or journalistic contexts. Homogenization is due to the fact that these models, which have been trained on an enormous variety of data, tend to reproduce what is statistically significant disproportionately, thus reinforcing biases and uniformity in generated results.</p><p>On the other hand, rights holders suffer from the fact that when their content is used as training data for AI models, they lose control over how this content will be used. Once ingested by AIs, this content will become part of the models’ imagination and can be regurgitated in different forms, more or less similar to their original form, without rights holders being able to either validate or prevent these uses. And obviously, without being able to obtain adequate remuneration for these uses, it is practically impossible to calculate precisely which training data contributed to enabling the creation of new content by AI.</p><p>The good news is that there exists a common solution element to all these problems: the solution consists of using complementary modules that allow enriching artificial intelligence models with specialized data (an artist’s works, or a singer’s voice), without directly integrating them into the base model. Instead of training the model with protected data, specialized circuits are put in place that plug into these models to provide them with more specific information in real-time. This allows AI solution developers to offer better performing services while ensuring that protected data remains isolated and separate from the model.</p><p>And these solutions already exist. This is what Mistral did with its partnership with AFP — enriching its language models with AFP dispatches to provide correct and verified information about current events. While offering remuneration to AFP.</p><h4>So how can we generalize this model into a commercial offer that could have the same impact that “streaming” had in the 2000s?</h4><p><a href="https://app.alias.studio/">Alias.studio</a> is a platform that allows artists to train models from their own creations and then make them available to the public under specific conditions. If I want to enrich my AI with <a href="https://app.alias.studio/artist/ORLAN">ORLAN’s works</a>, for example, I can now do so, but only if I respect the conditions she has defined, including pricing conditions for the use of these works. This means that on one side, ORLAN maintains control over her creations but also benefits from a passive income stream generated by her own AI model. And, on the other side, I benefit from a higher-end solution, with an AI enriched by ORLAN’s works, and with legal certainty that everything generated by this AI has been not only certified (in terms of quality) but also authorized by ORLAN (with a copyright license) — and I will therefore be very happy to pay for this higher quality service, knowing that ORLAN will be the main beneficiary of my royalties.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EWx0kZb5bQv8C0gwIFu2pA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HP4efYFxr40jyVlzlteX7Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*B0FwSKM9yZhm4-38_wTAVw.png" /><figcaption><a href="https://app.alias.studio/artist/ORLAN/alias?aliasId=6596a7eaaf1b6f88e6cbe4ba">Mayas by ORLAN on Alias</a></figcaption></figure><p>Beyond images and artists, the same principle applies to texts, music, or videos, as well as the voice or face of personalities. This protected data can be used to enrich AIs without losing control over how it is used.</p><p>Thus, as the Internet did in the past, AI raises many questions about copyright and somewhat forces the reinvention of cultural industries’ business models. But as with the Internet, it would be a mistake to consider AI only as a threat, forgetting all the opportunities it also brings. If streaming allowed the creation of new sources of passive income — both for creators and content publishers — generative AI is also a potential source of passive income, both for artists and cultural industries.</p><p>The important thing is to seize the opportunities of generative AI, but to seize them in a thoughtful and deliberate way. Artists, creators, and content publishers now have the opportunity to create their own models, with their own data, and make them available to all those who want to develop high-end AIs and provide superior quality services to traditional AIs, which suffer from hallucinations and homogenization. And this, while remaining in respect of copyright, ensuring control and fair remuneration for rights holders.</p><p>Because it is by enriching AIs (with quality data) that we can effectively enrich humanity (in both senses of the word).</p><p><strong>Follow Primavera de Filipi<br></strong><a href="https://x.com/yaoeo">X</a></p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><strong>Check out Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=45a4f5ed94b4" width="1" height="1" alt=""><hr><p><a href="https://papers.alien.club/what-is-the-equivalent-of-streaming-for-generative-ai-45a4f5ed94b4">What is the equivalent of streaming for generative AI?</a> was originally published in <a href="https://papers.alien.club/">alien.papers</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Explain Me Like I’m Five: Retrieval-Augmented Generation (RAG)]]></title>
            <link>https://papers.alien.club/explain-me-like-im-five-retrieval-augmented-generation-rag-11ebd9307ef7?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/11ebd9307ef7</guid>
            <category><![CDATA[rags]]></category>
            <category><![CDATA[retrieval-augmented-gen]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Tue, 18 Feb 2025 20:59:56 GMT</pubDate>
            <atom:updated>2025-02-18T20:59:56.391Z</atom:updated>
            <content:encoded><![CDATA[<p><em>by Primavera de Filippi</em></p><p>Imagine you’re playing with your favorite superhero action figure. Let’s call him AI-Man. AI-Man is really smart and can answer lots of questions, but sometimes he needs help remembering all the details — just like how you might need help remembering everything you learned in school.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*v0IynDVVJLtX_GrcEWQILg.png" /><figcaption>Image created with the help of <a href="https://app.alias.studio/artist/diminator/alias?aliasId=65cf9aa126e32ab15b2eb38a">Modica by Dimi by diminator on Alias</a></figcaption></figure><p>Now, imagine AI-Man has a magical library full of books. This isn’t just any library — it’s his personal helper that makes him even smarter! This is exactly how Retrieval-Augmented Generation (RAG) works. Let’s break it down with our superhero story.</p><h3>The Superhero and His Library</h3><p>When someone asks AI-Man a question, he doesn’t just make up answers from his basic training. Instead, he runs to his magical library and looks for books that might help him answer the question better. This is the “Retrieval” part — it’s like having a super-fast librarian who can find exactly the right books in seconds!</p><p>For example, if someone asks AI-Man about the latest company policies, he quickly searches through his library of company documents. Or if someone wants to know about customer feedback from last month, he checks his collection of customer reviews.</p><h3>How AI-Man Uses His Library</h3><p>Here’s the cool part: AI-Man doesn’t just read the books word for word to people. He’s much smarter than that! He:</p><ol><li>First, finds the most relevant books (retrieval)</li><li>Then, reads and understands the important parts</li><li>Finally, explains everything in his own words, combining his general knowledge with the specific information from the books</li></ol><p>This is the “Generation” part — he generates new, helpful answers based on both his training and the extra information from his library.</p><h3>Why This Makes AI-Man a Better Superhero</h3><p>Remember when you had to do a school presentation? You probably didn’t just use what you already knew — you looked things up in books or asked your parents for help. That’s exactly why RAG makes AI systems better:</p><ul><li>They’re more accurate because they use up-to-date information from their “library”</li><li>They can talk about specific details they wouldn’t know otherwise</li><li>They can prove where their information comes from by pointing to their sources</li><li>They’re less likely to make mistakes or invent false information</li></ul><h3>Real-World Super Powers</h3><p>In the real world, RAG systems are helping companies do amazing things:</p><ul><li>Customer service bots that can answer questions about specific products and their latest features</li><li>Research assistants that can analyze thousands of scientific papers in seconds</li><li>Technical support systems that always know the most recent troubleshooting steps</li><li>Documentation helpers that can explain complex systems using the most current information</li></ul><h3>When Things Get Tricky</h3><p>Sometimes, even superheroes face challenges. AI-Man’s library system isn’t perfect:</p><ul><li>He needs to make sure his books are well-organized</li><li>He has to be good at picking the right books for each question</li><li>His library needs regular updates to stay current</li><li>He needs to be careful not to mix up different pieces of information</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MnjZoqdTvO-QY03dQ4BLxg.png" /><figcaption>Image created with the help of <a href="https://app.alias.studio/artist/diminator/alias?aliasId=65cf9aa126e32ab15b2eb38a">Modica by Dimi by diminator on Alias</a></figcaption></figure><h3>The Future of Our Superhero</h3><p>As AI-Man gets better at using his library, he becomes an even more powerful superhero. Scientists are constantly helping him:</p><ul><li>Find books faster</li><li>Understand the contents better</li><li>Combine information more effectively</li><li>Keep his library organized and up-to-date</li></ul><p>So next time you hear about RAG, remember AI-Man and his magical library. It’s all about making AI systems smarter by giving them access to specific, relevant information when they need it — just like how you become smarter by learning from books, teachers, and experiences!</p><p><strong>Follow Primavera de Filipi<br></strong><a href="https://x.com/yaoeo">X</a></p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><strong>Check out Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=11ebd9307ef7" width="1" height="1" alt=""><hr><p><a href="https://papers.alien.club/explain-me-like-im-five-retrieval-augmented-generation-rag-11ebd9307ef7">Explain Me Like I’m Five: Retrieval-Augmented Generation (RAG)</a> was originally published in <a href="https://papers.alien.club/">alien.papers</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[DeepSeek and the AI copyright wars: The empire strikes back?]]></title>
            <link>https://papers.alien.club/deepseek-and-the-ai-copyright-wars-the-empire-strikes-back-71f90848e3cd?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/71f90848e3cd</guid>
            <category><![CDATA[ai-copyright]]></category>
            <category><![CDATA[deepseek]]></category>
            <category><![CDATA[china]]></category>
            <category><![CDATA[janus-pro]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Tue, 11 Feb 2025 10:12:12 GMT</pubDate>
            <atom:updated>2025-02-11T10:12:12.163Z</atom:updated>
            <content:encoded><![CDATA[<p><em>by Primavera de Filippi</em></p><p>DeepSeek is a Chinese AI company that has recently gained attention for its advanced AI models, particularly DeepSeek R1, which was released on January 20th 2025 — just one day before the announcement of the <a href="https://openai.com/index/announcing-the-stargate-project/">Stargate Project by OpenAI</a>, a new company which intends to invest $500 billion over the next four years to build new AI infrastructure for OpenAI in the United States. The release of R1 triggered a Silicon Valley panic attack. The reason is that R1 is not just another large language model (LLM) that merely predicts the next word based on patterns in text, it’s a reasoning model that has been specifically trained to achieve structured problem-solving and logical deduction, making it more akin to human reasoning. As of today, R1 rivals top reasoning models from OpenAI (e.g. o1) and Google (e.g. Gemini Flash Thinking), but at a fraction of the cost. <strong>Besides, it’s open source. Think ChatGPT with a sharper brain, a smaller waistline, and an MIT license stapled to its forehead.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iarzJ9yjoBFRy98sG9S_bw.png" /><figcaption>Image created using <a href="https://app.alias.studio/artist/Gregorio%20Undurraga">Undurraga Oil Paintings v0.70 on Alias</a></figcaption></figure><p>Importantly, DeepSeek’s release of R1 isn’t just a matter of geopolitics and market competition, it’s also a legal earthquake that could reshape AI regulation worldwide.</p><h3>A copyright nightmare in the making</h3><p>The first red flag is the training data. <a href="https://github.com/deepseek-ai/DeepSeek-R1/blob/main/DeepSeek_R1.pdf">DeepSeek R1 research paper</a> doesn’t spell out exactly what it was trained on — which is not surprising, as this is a common practice in the generative AI landscape not to disclose the full extent of the training data set. Hence, we can only attempt to infer R1’s training dataset by looking at the training data of previous generative models developed by DeepSeek. Former LLMs were trained on CommonCrawl, whereas DeepSeek vision-language model (VLM) was trained, amongst other things, on a variety of repositories of unlicensed copyright works, such as LibGen, Sci-Hub and other sources of pirated papers and books. While there is no proof that R1 was trained on these ‘shadow libraries’, if it were, it would potentially bring a case for copyright infringement.</p><p>Meta already got blamed for <a href="https://www.theverge.com/2025/1/14/24343692/meta-lawsuit-copyright-lawsuit-llama-libgen">allegedly using LibGen</a> as part of LLama’s training data, and <a href="https://www.hollywoodreporter.com/business/business-news/openai-training-data-inspected-authors-copyright-case-1236011291/">OpenAI is already fighting multiple lawsuits</a> over its training practices. But here’s the twist: DeepSeek is based in China. Suing them for copyright infringement outside of China might be a challenge in and of itself. Besides, even if someone could build a case for copyright infringement, proving direct damage from AI training is still an uphill battle. If DeepSeek’s model doesn’t spit out copyrighted books verbatim, courts might struggle to see any harm done. In short, it’s the Wild West — except China gets to write the law.</p><h3>Did DeepSeek “steal” from OpenAI?</h3><p><a href="https://www.404media.co/openai-furious-deepseek-might-have-stolen-all-the-data-openai-stole-from-us/">OpenAI claims that DeepSeek engaged in model distillation</a> — a process where you prompt another generative AI model with many questions, record its answers, and use that as training data for your own model. Think of it as cloning a Michelin-starred chef’s recipes by eating at their restaurant every night in order to learn to replicate their secret recipes. If true, this means DeepSeek effectively piggy-backed on OpenAI’s $500 million R&amp;D budget to launch a competing model at a fraction of the cost.</p><p>But is this copyright infringement? Probably not. OpenAI might be able to argue a <strong>terms of service</strong> violation if DeepSeek scrapped its API, but that’s a breach of contract, not a copyright claim. And enforcing terms of service across international borders is another legal headache. Even if OpenAI won a lawsuit, it’s unclear how they’d stop DeepSeek from training another model using the same trick.</p><h3>The future of AI copyright battles</h3><p>DeepSeek’s emergence isn’t just another chapter in the AI copyright saga — it’s a plot twist. It forces us to rethink who controls AI training data, whether copyright law is enforceable at scale, and whether proprietary AI can survive against open-source alternatives.</p><p><strong>1. Shadow libraries: pragmatism vs. copyright law</strong></p><p>If DeepSeek R1’s strong performance is linked to training on shadow libraries like LibGen and Sci-Hub, this could <strong>set a dangerous precedent</strong> — or at least an unavoidable one. AI developers will take note: if pirated data improves AI reasoning, it’s only a matter of time before others follow suit, whether openly or in secret.</p><p>This creates a direct collision between copyright law and market pragmatism. AI companies aren’t driven by ethics; they’re driven by performance. If the best models emerge from unlicensed data, market forces will push developers toward that path — whether through plausible deniability, legal loopholes, or outright defiance.</p><p>Legally, enforcement is complicated. If DeepSeek operates primarily in China, Western copyright lawsuits won’t touch it. But what happens when open-source models trained on this data start circulating globally? Courts in the U.S. and EU will face a dilemma: either try to block AI models trained on unauthorized data (good luck policing that) or accept a new reality where AI is built on imperfect, legally gray foundations.</p><p><strong>2. AI arms race: when geopolitics overrides copyright</strong></p><p>The AI race isn’t just about technology — it’s about <strong>national power</strong>. The U.S., China, and the EU all see AI as a matter of economic and military dominance. In that context, copyright enforcement becomes a secondary concern at best.</p><p>The U.S. won’t let copyright lawsuits cripple domestic AI development if China is forging ahead. If AI models trained on copyrighted material prove significantly more capable, policymakers might push for legal carve-outs to shield AI companies from copyright liability, just as they’ve done for tech firms in the past (e.g., DMCA safe harbors, Section 230 for platforms).</p><p>The EU, meanwhile, has taken a stricter approach with its AI Act, which could mean that the EU is doomed to fall behind. China, for its part, has no intention of slowing down — meaning global AI regulation may eventually bend toward competitiveness, not copyright purity.</p><p><strong>3. Open Source vs. closed AI models: a new competitive pressure</strong></p><p>DeepSeek’s open-source release of its reasoning model creates a strategic dilemma for proprietary AI companies like OpenAI, Google, and Anthropic. They’ve built their business models on closed, high-cost development, assuming their models would always be superior to open-source alternatives. But what if they aren’t?</p><p>If open-source AI can achieve <strong>comparable or better</strong> performance, the closed-model approach faces the existential threat of market disruption. Why pay OpenAI or Google for access when you can run a state-of-the-art model locally for free?</p><p>Besides, the entire premise of copyright lawsuits against proprietary models may collapse if open-source AI models are developed by community-driven innovation. If the best AI models are open and free, copyright fights over training data may become a footnote in history.</p><h3>Conclusion: A copyright reckoning is coming</h3><p>DeepSeek release of R1 raises serious questions about copyright enforcement, data ownership, and whether <strong>copyright law is even relevant in an AI-dominated world</strong>. If copyright is supposed to incentivize human creativity, what happens when the most advanced “creators” are machines that remix and reassemble information faster than we can legislate?</p><p>DeepSeek R1 isn’t just another AI model — it’s a <strong>stress test for AI copyright law</strong>. If shadow libraries prove too valuable to ignore, if national AI policy overrides copyright concerns, and if open-source AI outpaces proprietary models, then the copyright battles we’re seeing today may soon feel outdated, futile, or both.</p><p>Ultimately, if DeepSeek can prove anything, it is that the age of AI copyright wars is just beginning, and no one — neither the corporations nor the regulators — knows how this story ends. The fundamental question isn’t whether AI will change copyright — it’s whether copyright is capable of withstanding AI.</p><p><strong>Follow Primavera de Filipi<br></strong><a href="https://x.com/yaoeo">X</a></p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><strong>Check out Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=71f90848e3cd" width="1" height="1" alt=""><hr><p><a href="https://papers.alien.club/deepseek-and-the-ai-copyright-wars-the-empire-strikes-back-71f90848e3cd">DeepSeek and the AI copyright wars: The empire strikes back?</a> was originally published in <a href="https://papers.alien.club/">alien.papers</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Action II Workshop at Chateau du Fey]]></title>
            <link>https://medium.com/@alien.papers/ai-action-ii-workshop-at-chateau-du-fey-89cde086ccf8?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/89cde086ccf8</guid>
            <category><![CDATA[ai-governance]]></category>
            <category><![CDATA[ai-rights-management]]></category>
            <category><![CDATA[alien-intelligence]]></category>
            <category><![CDATA[chateau-du-fey]]></category>
            <category><![CDATA[action-ai]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Mon, 10 Feb 2025 18:59:09 GMT</pubDate>
            <atom:updated>2025-02-10T18:59:09.001Z</atom:updated>
            <content:encoded><![CDATA[<p>Following the AI Action Summit in Paris, the upcoming <strong>AI Action II: Decompression in the Latent Space</strong> workshop at Chateau du Fey (February 12–15, 2025) marks a pivotal moment in the evolution of AI rights management (AI-RM), knowledge preservation, and digital governance. This gathering, set in the historic Burgundy countryside, brings together an unparalleled roster of experts across technology, academia, law, and civil society to chart the future of responsible AI development.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eHkRG6aVGCKj8cZHnko6CQ.png" /><figcaption>Image created using <a href="https://app.alias.studio/artist/Gregorio%20Undurraga">Undurraga Oil Paintings v0.70 on Alias</a></figcaption></figure><h3>AI Action II: a landmark event for AI-RM</h3><p>AI-RM stands at the heart of critical conversations surrounding AI training, knowledge access, and digital ownership. As AI systems increasingly shape our digital landscape, the need for robust frameworks that respect rights while fostering innovation has never been greater. AI Action II is a <strong>hans-on solutions-oriented workshop</strong> that dives into pressing challenges such as:</p><ul><li><strong>Access to Knowledge in the AI Age</strong> — How can we manage AI training data responsibly? What rights frameworks can ensure ethical curation of knowledge bases?</li><li><strong>AI &amp; World Libraries</strong> — The transformation of global knowledge infrastructure through AI, exploring new forms of archives, provenance tracking, and credibility mechanisms.</li><li><strong>Small is Beautiful: Narrow AI &amp; Public Benchmarks</strong> — Beyond large-scale AI models, how can we promote specialized, localized AI models that serve diverse communities?</li><li><strong>Agentic AI</strong> — The implications of AI autonomy, governance structures for AI agents, and evolving concepts of authorship and rights.</li></ul><p>With an emphasis on <strong>practical outcomes</strong>, the workshop will address emerging legal standards for generative AI licensing and certification, as well as the role of Retrieval-Augmented Generation (RAG) systems for upgrading generative AI models, and their integration into decentralized scientific publishing.</p><h3>The Role of Alien Intelligence in Spearheading AI-RM</h3><p><strong>Alien Intelligence </strong>is at the forefront of AI-RM, driving critical conversations on legal tech solutions and rights-based AI licensing and certification. By bringing together diverse stakeholders — from AI developers to legal scholars — this workshop fosters <strong>cross-disciplinary collaboration</strong> to build sustainable AI knowledge ecosystems.</p><p>Through its involvement in the organisation of this workshop, <strong>Alien Intelligence is shaping the global legal and knowledge infrastructure for AI</strong>, ensuring that AI’s development aligns with principles of trust, transparency, and accountability.</p><h3>A Gathering of Visionaries</h3><p>AI Action II is more than just a workshop — it’s a <strong>defining moment for AI rights management and AI governance</strong>. By convening leading voices in an intimate, collaborative environment, this event will lay the groundwork for innovative approaches to AI development that respect knowledge, rights, and human agency.</p><p>The workshop benefits from an extraordinary lineup of participants, including <strong>Brewster Kahle (Internet Archive), EM Lewis-Jong (CommonVoice), Sonam Jindal (Partnership on AI)</strong>, <strong>Pedro Ortiz Suarez (Common Crawl) </strong>and representatives from leading organizations such as the <strong>Aspen Institute, Internet Archive </strong>and<strong> Mozilla Foundation</strong>. Co-organized by <strong>Primavera De Filippi </strong>and<strong> Samuel J. Klein</strong>, the workshop builds on the success of previous AI Renaissance workshops in Germany, continuing the tradition of fostering groundbreaking discussions in an immersive setting.</p><p>Stay tuned for updates, insights, and post-event reflections as we continue to push the boundaries of AI right management and knowledge preservation.</p><p><strong>Follow Alien<br></strong><a href="https://discord.com/invite/TpTqgZgDwN">Discord Community<br></a><a href="https://www.linkedin.com/company/alienintelligence">Linkedin<br></a><a href="https://www.instagram.com/aliasbyalien/">Instagram<br></a><a href="https://x.com/alienintelai">X</a></p><p><strong>Check out Alias Studio</strong><a href="https://www.alias.studio/"><br>Website</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=89cde086ccf8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Alien’s collaboration with Stanford CodeX Blockchain Group]]></title>
            <link>https://medium.com/@alien.papers/aliens-collaboration-with-stanford-codex-blockchain-group-1bd4ecfd4f61?source=rss-e5b0a7c9d40d------2</link>
            <guid isPermaLink="false">https://medium.com/p/1bd4ecfd4f61</guid>
            <category><![CDATA[intellectual-property]]></category>
            <category><![CDATA[generative-ai]]></category>
            <category><![CDATA[ai-governance]]></category>
            <category><![CDATA[blockchain]]></category>
            <dc:creator><![CDATA[Alien Intelligence]]></dc:creator>
            <pubDate>Fri, 31 Jan 2025 12:31:45 GMT</pubDate>
            <atom:updated>2025-01-31T12:31:45.460Z</atom:updated>
            <content:encoded><![CDATA[<h3>A collaborative effort to design and promote blockchain-based IP tools and standards for rightful generative AI</h3><p><em>by Tony Lai &amp; Primavera de Filippi</em></p><p>The exponential growth of generative AI technologies has transformed how we produce, share, and monetize digital content, creating a complex landscape that simultaneously threatens and empowers creative professionals. Indeed, while generative AI presents significant challenges to the economic livelihood of many artists and creators, it also presents many new opportunities for creative expression and collaborative production — ultimately inviting us to re-conceptualise our relationships to creativity, authorship, and artistic production.</p><p>Artists are beginning to view AI not as a replacement, but as a collaborative tool — a new medium that can extend and augment human creativity. It is allowing them to prototype and iterate through designs more rapidly, as well as to experiment with new forms of artistic expression that blend human creativity with machine-generated outputs. Generative AI challenges us to reimagine creativity not as a purely individual act, but as a more fluid, networked form of collaborative production that can transcend traditional boundaries between human and machine, original and derivative, creation and interpretation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*j_PlQ4EQDDZ9ccBi" /><figcaption><em>An image generated through </em><a href="https://app.alias.studio/artist/primavera?utm_source=landing"><em>author of this article ’s Alias</em></a><em> on alias.studio, a platform for artists to create generative ai models, generate revenue and protect their creative rights.</em></figcaption></figure><p>This, of course, also requires new Intellectual Property (IP) management frameworks. A variety of regulatory reforms are on the way (e.g. the AI Act in Europe and the California AI Transparency Act in the U.S.) but are currently limited and fragmented in scope. Existing regulations focus on narrow aspects like data protection, algorithmic transparency, or copyright and data mining — reflecting a piecemeal approach that fails to address the comprehensive challenges posed by generative AI. Key limitations in current regulatory approaches include insufficient mechanisms for tracking and attributing AI-generated content and inadequate frameworks for compensating original creators whose work trains AI models.</p><p>The regulatory challenges of generative AI mirror historical precedents, particularly the early internet era when existing legal frameworks struggled to address the new practices of digital content sharing. Peer-to-peer file-sharing once questioned the viability of traditional copyright enforcement. Similarly, generative AI requires a fundamentally new approach to IP management and creative production. To address these challenges, we need flexible IP frameworks that recognize the collaborative nature of AI-assisted creation, along with new mechanisms for transparent attribution and fair compensation. As the law is struggling to cope with the rapid pace of innovation in the generative AI industry, we believe that new “private ordering” mechanisms are required to quickly adapt to these technological innovations.</p><p><strong>A journey towards a consolidated, holistic approach</strong></p><p>At the <a href="https://law.stanford.edu/projects/blockchain-group/">CodeX Blockchain Group</a>, we research and publish informed perspectives around the blockchain ecosystem and seek to become an inclusive, neutral forum for regulators and policy-makers to convene with academics, professionals, and technologists across Stanford and the wider blockchain community. We see blockchain-based systems as expressions of computational law, with many unique properties, including around transparency, interoperability and incentivization.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ww5Ylo5yRacHbsswBSdaNA.png" /><figcaption><em>Traditional vs. Blockchain-enabled IP Management</em></figcaption></figure><p>Alongside our wider community of collaborators and partners, we have been gathering across various conferences, projects, jurisdictions and working groups to address some of the generative AI opportunities and challenges by leveraging blockchain technology to design new, interoperable computational law standards for managing IP rights. In expanding this work to the design of new open, decentralized, and transparent standards for generative AI, we aim to establish a foundation for interoperable IP management frameworks that can meet the needs of new forms of collaborative activities — from art, to business, to civic participation and democratic dialogue. These standards include the embedding of legal licenses into NFTs (Non-Fungible Tokens), providing a transparent, automated solution for IP licensing, fostering innovation while protecting the rights of creators and stakeholders.</p><p>For example, imagine a visual artist who licenses their work for training a generative AI model. Without standardized IP frameworks, the artist risks losing control over how their work is used or credited. Using blockchain-enabled NFTs with embedded legal licenses, the artist could specify terms for how their artwork can be used in training or incorporated into outputs, or track and verify where and how their work is used, ensuring attribution and fair compensation. As a result, a generative AI platform could automatically enforce royalty payments whenever the platform generates an image inspired by the artist’s original work.</p><p>We envision a world where artists, businesses, civic actors, and developers can collaborate seamlessly, while generating and respecting IP rights. At the core of this vision is the development of a standardized governance framework that promotes collaboration amongst the various stakeholders of the generative AI industry that are committed to the “rightful use” of AI.</p><p>This includes the creation of a collaborative endeavor, leveraging IP law (trademark and copyright) to promote collaboration, innovation, and safety in the AI ecosystem. At the heart of this initiative is a publicly licensed trademark that signals compliance with ethical and legal standards for generative AI applications, ensuring a recognized and trusted brand for all participating entities. This trademark can be tied to a charter that serves as its licensing mechanism and outlines the principles and commitments required for joining the collaborative endeavor. Members can self-select to use the label, but if they do so, they must legally commit to abiding by the rules of the associated charter. Importantly, by requiring members to provide privileged access conditions to their IP assets to all other members of the initiative (e.g. through a Copyfair license), we aim to move away from the traditional payoff structure whereby competition is the dominant strategy towards a new payoff structure that incentivizes cooperation, fostering a more collaborative and innovative environment. This is particularly relevant for the generative AI ecosystem, where it has become extremely difficult for emergent players and startups to compete on a level playing field with the established dominant players.</p><p>We are grateful to every individual and group we have engaged with around this vision, who each offer unique insights and expertise for the journey ahead: Megan Ma, Roland Vogl, and the CodeX community offer established leadership in computational law, and insights into intersections with legal education, I.P., and A.I. The team at <a href="https://alien.club/">Alien</a> develops blockchain-based legal technologies, particularly token-bound licenses, to facilitate transparent and automated IP licensing. We work alongside Josh Tan, B Cavallo and other researchers from the <a href="https://metagov.org/">Metagov</a> and <a href="https://publicai.network/">Public AI</a> ecosystems, curating discussions and sharing ideas at the frontier of AI governance and decentralized systems. With <a href="https://www.radicalxchange.org/">RadicalxChange</a> we focus on reimagining social institutions, and through their Legal Design Lab have been working on innovative approaches to governance and legal design through systems of private law. The team at <a href="https://block.science/">Blockscience</a> with their proficiency in complex systems engineering and blockchain analytics, assist in designing and analyzing the technical aspects of the proposed IP management systems, ensuring they are robust, scalable, and effective. As an artist and filmmaker integrating AI into creative works, <a href="https://www.lawrencelek.com/">Lawrence Lek</a> offers a unique perspective on the practical implications of IP management in AI-generated art, contributing to the development of standards that protect and promote creative expression. Dedicated to building tools for rights holders and AI developers, Matt Dryhust, Holly Herndon, Jordan Meyer, Cullen Miller, and the team at <a href="https://spawning.ai/">Spawning</a> share their expertise in creating ecosystems that respect and protect IP rights, aiding in the development of tools and standards that prevent unauthorized use of protected materials in AI training. This builds upon the work done by <a href="https://openfuture.eu/">Open Futures</a>, led by Paul Keller and Alek Tarwokwsi, focused on promoting the digital commons and fostering better policies for right holders in an open Internet. As a nonprofit R&amp;D lab focused on building AI tools for coordination and deliberation, Colleen McKenzie and the team at the <a href="https://ai.objectives.institute/">AI Objectives Institute</a> help our work be informed by cutting-edge AI capabilities and aligned with broader societal objectives. Through our discussions with Tenzin Seldon, an advisor to the <a href="https://hai.stanford.edu/">Stanford Human-Centered AI project</a>, we benefit from her ongoing insights into ethical AI development and deployment.</p><p>As we continue our journey alongside these and many other allied friends and colleagues, we are forming a dedicated task force of legal scholars, technologists, and IP stakeholders to collaboratively design, test, and refine the framework. We hope this task force can channel this wealth of expertise spanning computational law, blockchain systems, decentralized governance, and creative industries, to nurture frameworks for collaboration that will help drive innovation, inclusivity, and accountability across the whole lifecycle of Generative AI, from model training to content generation, and in a diverse array of public and civic interest use cases, from art and creativity to health and climate action, from circular economies and community currencies, to food systems and regenerative agriculture.</p><p>In the coming months, we aim to:</p><ul><li>Formalize partnerships and finalize criteria for participation in the task force.</li><li>Develop a research roadmap with clear milestones and deliverables.</li><li>Launch pilot projects to test and validate the framework.</li><li>Engage with IP lawyers, creators, and industry leaders through workshops and at CodeX FutureLaw, in April 2025.</li></ul><p><strong>Join Us</strong></p><p>We invite researchers, technologists, creators, and policymakers to join us in shaping the future of Generative AI and IP standards, including at <a href="https://conferences.law.stanford.edu/futurelaw2025/">CodeX FutureLaw on 10 April 2025</a>, and in our CodeX Blockchain Group activities alongside. Together, we can create a transparent, fair, and collaborative ecosystem for the rightful use of AI.</p><p><strong>Tony Lai </strong>is a Fellow at CodeX, the Stanford Center for Legal Informatics, leads the <a href="https://law.stanford.edu/projects/blockchain-group/">CodeX Blockchain Group</a>, and is a Founding Editor of the <a href="https://stanford-jblp.pubpub.org/about">Stanford Journal of Blockchain Law &amp; Policy</a>. He is a co-founder of legal-tech company <a href="http://legal.io/">Legal.io</a>, and of the non-profit research organization, <a href="https://www.mothertreelabs.com/">Mothertree Labs</a>. He specializes in governance tooling for public goods and ecological transition.</p><p><strong>Primavera de Filippi</strong> is a legal scholar at Harvard and cofounder of <a href="https://alien.club/">Alien</a>. She specializes in legal-tech solutions for AI, and has been developing blockchain-based legal technologies, particularly token-bound licenses, to facilitate transparent and automated IP licensing.</p><p>For inquiries and updates, please <a href="/service/mailto:tony@codex.stanford.edu?cc=pdefilippi@gmail.com&amp;subject=Inquiry%3A%20COLLAIPA">contact the CodeX Blockchain Group</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1bd4ecfd4f61" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>