I don’t think this is a Marxist perspective. It is founded on an assumption that LLMs are capable of positively transforming creative production in one way or another, be it by raising productivity of existing participants or lowering barriers of entry for new ones. This assumption is false, in my opinion, because there is a difference between abundance and dilution. When the market is flooded with lemons of questionable quality, all that is achieved is intensified information asymmetry, meaning that it becomes harder, more labourious, more expensive to distinguish good lemons from bad ones. This leads to a less productive market overall because sellers have to invest extra labour and money to even have a chance to find buyers while buyers invest the extra to buy or produce information that would help discover the good lemons. A good existing example of this situation is the today’s job market.
And if one is to argue that LLM produced slop is the good lemon – or going to be one very soon just you wait Sam Altman / China is developing it so fast! – then they arrive at, in my opinion, the crux of the negative reaction towards the LLM hype. It is not any of the reasons listed in the OP, it is just that people think it is not very useful or good.
I don’t think this is a Marxist perspective. It is founded on an assumption that LLMs are capable of positively transforming creative production in one way or another, be it by raising productivity of existing participants or lowering barriers of entry for new ones. This assumption is false, in my opinion, because there is a difference between abundance and dilution. When the market is flooded with lemons of questionable quality, all that is achieved is intensified information asymmetry, meaning that it becomes harder, more labourious, more expensive to distinguish good lemons from bad ones. This leads to a less productive market overall because sellers have to invest extra labour and money to even have a chance to find buyers while buyers invest the extra to buy or produce information that would help discover the good lemons. A good existing example of this situation is the today’s job market.
And if one is to argue that LLM produced slop is the good lemon – or going to be one very soon just you wait Sam Altman / China is developing it so fast! – then they arrive at, in my opinion, the crux of the negative reaction towards the LLM hype. It is not any of the reasons listed in the OP, it is just that people think it is not very useful or good.