AI and the Destruction of the Creative Commons

Copyright and openness have long been a delicate balance in software protection. Over four decades of debate, an equilibrium was eventually reached. However, the emergence of artificial intelligence has shattered that equilibrium. As a young child, I recall typing BASIC programs from magazines into my Commodore 64 and later learning REXX to craft games for a BBS I operated.

These "open" examples were instrumental in my education, teaching me the fundamentals of programming. This period coincided with ongoing discussions about the copyright status of software. Initially, shareware and freeware were prevalent. Shareware was essentially trial-ware, allowing users to test the software and pay a small fee to unlock the complete version.

Freeware, on the other hand, was entirely free, but the source code was not disclosed. Doom, for instance, was distributed as shareware, generating a substantial market through word-of-mouth sharing. In response to the growing openness, copyright laws adapted, giving rise to "copyleft" licenses such as GPL, MPL, and CC-SA. These licenses compelled individuals who sought the benefits of free and open-source software to extend those privileges to any subsequent derivative works.

The advent of the modern internet and cloud infrastructure could not have been possible without free and open-source software. Every cloud service and the internet backbone itself are derivative works built upon the foundation laid by previous generations. Without such openness, accessing information online would be akin to using AOL or CompuServe, would be significantly more expensive, and would result in an even more monopolistic digital landscape.

While there is considerable debate surrounding the environmental impact, cybersecurity concerns, and the spread of misinformation due to generative AI large language models, the destruction of our foundational openness has received less attention. These AI models indiscriminately consume all online content, disregarding copyright or licensing restrictions.

Consequently, derivative works may or may not adhere to the licenses of their original creators, and there seems to be no willingness to enforce these obligations legally. This shift has led to a situation where I am more inclined to withhold my work while also being cautious about consuming information shared online. The presence of "slop code" or malicious libraries in shared content introduces additional risks, such as the potential for third-party data exfiltration or incorporation of stolen works, implicating the creator in criminal activities.

Conversely, if I make my code publicly available, AI tools can more easily identify vulnerabilities, enabling malicious exploitation of coding errors. Conversely, keeping my code closed offers greater protection against these threats. Moreover, publishing code on platforms like GitHub exposes it to a deluge of pull requests generated by AI bots and novice users, consuming significant time and resources in assessing their relevance.

Our personal ethical and moral principles are also diluted when absorbed into the vast AI ecosystem. Existing licenses and agreements, although imperfect, have been proven legally enforceable, granting creators control over their creations. However, AI has transformed the act of sharing knowledge from a generous gift to the world into a liability for authors.

In the era of AI, sharing knowledge risks becoming a liability for creators, undermining the power of openness and potentially leading to a consolidation of wealth and power. This shift has the potential to have far-reaching negative consequences for the future. The past 35 years of openness have generated nearly everything we benefit from online, fundamentally transforming the world in predominantly positive and empowering ways.

As we enter what could be likened to a digital dark age, it is crucial for authors, coders, technologists, and artists to unite and advocate for a new digital renaissance before significant losses occur. Failure to do so may result in a dystopian scenario reminiscent of George Orwell's "Fahrenheit 451."