Nothing has changed about the dangers of large language models but that doesn't stop me from enjoying the tech bros meltdowns over DeepSeek. Drinking their own Kool-Aid and then trying to come up with all sorts of explanations for why no "the Chinese aren't innovating."