It's easy to miss his real point due to the provocative headline - the froth and churn of AI in 2022-2024 is full of con artistry, but that doesn't mean it's al

It's easy to miss his real point due to the provocative headline - the froth and churn of AI in 2022-2024 is full of con artistry, but that doesn't mean it's all useless: "But it's not all doom and gloom. Doctorow believes that the situation isn't a zero-sum game, and that once the dust settles, there may be some residual upshots that could drive meaningful technological progress in the future. 'Tech bubbles come in two varieties: The ones that leave something behind, and the ones that leave nothing behind,' Doctorow argued. Following the dotcom bubble popping, millions of young people were lured into the tech sector, creating an 'army of technologists.'" Transformer neural networks have been catastrophically overhyped - they are considerably dumber than bees - but they are clearly useful for lots of things, and the engineering work on getting them to run cheaply will pay dividends when LLMs are run on more advanced architectures.