The author argues that using copyright to prevent AI from training on content is short-sighted and could lead to a reduction in fair use protections. They suggest that this could lead to a situation where big companies can navigate the complex rights necessary to build a model, but new companies or open source models would be unable to do so. The author concludes by advocating for a more nuanced approach to copyright policy that addresses the whole problem, rather than using copyright as a "sledgehammer".
Key takeaways:
- Fair use allows people limited use of copyrighted material without permission and is crucial for many creators and industries.
- Some public interest advocates want to use copyright to stop AI from being trained on content without permission, which could weaken or destroy fair use.
- Using AI to analyze content, similar to the controversial "on a computer" software patents, is becoming a contentious issue in copyright policy.
- IP maximalization could lead to the creation of oligopolies, as big companies will be able to navigate the complex rights necessary to build a model, while new companies or open source models may struggle.