Oxford's Michael Bronstein disputes Gary Marcus's claim that generative AI is fizzling, arguing a two percent dip is not a systemic failure

- Marcus's Substack post argues AI hype outpaces actual performance.