AI has become as a deeply polarizing issue on the left, with many people having concerns regarding its reliance on unauthorized training data, displacement of workers, lack of creativity, and environmental costs.
It’s important to make a distinction between the definition of “open source AI” canonized by the OSI that doesn’t require open training data, and models where all of the training data used is also made available.
Separately, the tools most people think about when they hear “AI”, generalized generative AI models, only exist as capitalist surplus, and we shouldn’t be defending them. Hyper focused AI tools such as the Te Hiku Media project to create speech recognition tools for the te reo Māori language are unequivocally good, and we should be making a lot more projects like this.
It’s important to make a distinction between the definition of “open source AI” canonized by the OSI that doesn’t require open training data, and models where all of the training data used is also made available.
Separately, the tools most people think about when they hear “AI”, generalized generative AI models, only exist as capitalist surplus, and we shouldn’t be defending them. Hyper focused AI tools such as the Te Hiku Media project to create speech recognition tools for the te reo Māori language are unequivocally good, and we should be making a lot more projects like this.