GM Readers! ☀️

Welcome to the 20th issue of Evolving Internet Insights’ 🧠 Brainfood — a weekly deep dive into a relevant emerging tech topic.

While generative (gen)AI democratizes creation at the individual level, it also raises eyebrows around attribution and authorship. This is an area we have thought a lot about particularly as genAI adoption accelerates.

Today, we highlight how attribution and authorship has evolved over different eras of content creation, and why authorship and attribution is more challenging in the Age of genAI.

Thanks for reading!

Liang and Dan 🙌

🧠 Brain Food

One focus topic to feed your brain.

Time To Set Some Context 📚

GenAI technologies have drastically reshaped how content is created. An individual today can effortlessly create art, story lines, or melodies that would have historically required considerable time and resources. Prior to genAI, for example, creating a 4K ultra-realistic art piece required a team of highly skilled artists. With AI, such a feat is attainable by a single individual leveraging generative AI tools and a few prompts. AI truly democratizes content creation and unlocks a new wave of creative potential.

Yet, this newfound power to create, remix, repurpose, and reuse easier than ever before brings about new challenges around content authenticity–namely around attribution and authorship. 

At the core, it is about answering “who” created a piece of content. This seemingly simple question becomes increasingly harder to answer in a world of infinite creation ability enabled by AI.

Attribution Challenges Across Content Creation Eras 🖨

Over the course of history, each era of content creation introduced the need for a new paradigm for attribution.

During the pre-literary times, when content was primarily spoken, attribution and authorship was less of an issue. A skillful orator around a campfire got credit for the storytelling as the story was being told–it was more obvious who the author was.

As new technologies were introduced that enabled: (1) the replication of content, and (2) greater distribution of that content, the attribution challenge increased–who created what content became harder to answer. Across each era, new systems emerged to solve this problem.

With the invention of writing systems and their physical vessels (from papyrus to paper), content that was spoken could now be expressed on paper. This increased the risk of attribution as anyone can write down what a speaker said and disseminate it as their own. To get around this problem, signatures and seals were used to represent authorship.

The introduction of the printing press (e.g. the Gutenberg Press) accelerated the replication and distribution of content by enabling mass production (read: distribution). To enforce authorship and prevent others from stealing a creator’s works, intellectual property systems emerged.

The first few eras of content creation took place in the physical world. The Internet enabled infinite replication–a property that is unique to the digital world. We like to think the Internet is one big giant “copy and paste” machine. When you see a piece of content on the Internet, you can easily copy it and share it elsewhere, unlocking creation and distribution at a global scale. In this era, most of the attribution systems are enabled by modern content platforms. For example, when you publish a post on Linkedin, others can see you are the author who published that particular instance of the post. However, this system says nothing about whether you copied the idea or text from someone else. And even if you were the original author of that post, how does someone know you did not copy it from elsewhere? After all, all these systems are siloed walled gardens (see HBR article that Liang and Scott Kominers wrote) in that they do not “talk to each other.”

In short, with an increasingly digital native world on the Internet, attribution gets more challenging. Factors like the scale of content replication and distribution are foundational to this challenge, but genAI certainly further exacerbates it as well. With genAI, machines can now produce content that is influenced and inspired by their training data.

If someone uses genAI to create an image that looks like another image… who gets credit for it? What are the repercussions of this? How should the economic value (if any) be shared?

These issues are already coming to the forefront, with the Hollywood Writers and Actors strikes in 2023 and the recent NY Times lawsuit against OpenAI.

The Importance of Attribution

While attribution in the Age of genAI is an important problem, it can be difficult to grasp the magnitude of this problem. When people see the word “content,” they often think blogs, videos and content from the standpoint of entertainment and media. But anything in the knowledge work space can also be categorized as “content.” 

ARM, the AI chips design company, is a ~$70B company and has built its entire business on “content,” selling its designs and knowledge without manufacturing a single chip. All consulting firms also create “content”–when a consulting firm delivers a strategy for a client that is backed by research and analysis, that is “content”, albeit proprietary. We would go as far to say anything in the digital world can be viewed as “content.” With this definition, you can now see the implications of attribution in the Age of genAI.

Attribution and authorship serve as the underlying “accounting system” to reward those who create value. 

Here are some simple examples. Imagine the best selling book in the world, if we couldn’t determine who the author is, who would we pay? Who gets to capture the value this book has created? Or imagine the most useful piece of software in the world, if we do not know which company created it, who gets to capture that value? And value here doesn’t have to be monetary, it could also be non-monetary things like recognition. 

Would we want to create, innovate, and/or invent if there was no system of attribution? 

Would creators (individuals and companies) want to work extra hard to create something unique if someone else can easily take credit for the idea once it’s launched?

Ultimately, this is about “giving credit where credit is due.” An act that is becoming increasingly more challenging in the Age of genAI.

✍️📝

Readings:

• Threads Foreshadows a Big — and Surprising — Shift in Social Media: Liang and Scott Kominers wrote this Harvard Business Review piece explaining how social media sits in walled gardens and what an open source model would mean for content creators.

• Data Dignity: Developers Must Solve the AI Attribution Problem: an opinion piece on how the developer community can help fix the AI attribution problem with metadata solutions.

🗳 We’d Love your Feedback

🙏 Shameless Asks

It takes us days to put this together, sharing it takes you 19.45 seconds 😘

An easy way to support us and the newsletter is through the referral program.

⚡️ Share Evolving Internet Insights with a friend and ask them to subscribe

⚡️ Share on Twitter and Linkedin with a short note

⚡️ Share on your company Slack/Teams channels and communities

DISCLAIMER: This post is provided strictly for educational and informational purposes only. Nothing written in this post should be taken as financial advice or advice of any kind. The content of this post are the opinions of the authors and not representative of other parties. Empower yourself, DYOR (do your own research).