The article further explores the definition of AI and the concept of 'openness' in this context, arguing that it is often muddled and unclear. It also examines the components of AI systems, the ecosystem around open AI, and the dominance of corporate actors in the AI landscape. The article concludes by suggesting that the rhetoric of openness is often used in ways that exacerbate the concentration of power in the AI sector, rather than alleviating it.
Key takeaways:
- The article discusses the concept of 'open' artificial intelligence (AI), arguing that concepts from open-source software are being applied inappropriately to AI systems.
- Open AI systems can offer transparency, reusability and extensibility, but the rhetoric of openness is often used in ways that exacerbate the concentration of power in the AI sector.
- The definition of AI itself is contested and unclear, further complicating the question of what ‘open’ means in the context of AI.
- The article argues that the resources required to produce large-scale AI systems are concentrated in the hands of a few large tech companies, which not only shape what gets built but also the conditions under which AI systems can be built.