> Even if I’m wrong about that, I’m very confident that low quality, biased, and flat out incorrect web content was already a problem before LLMs. Definitely,

> Even if I’m wrong about that, I’m very confident that low quality, biased, and flat out incorrect web content was already a problem before LLMs. Definitely, and I believe the post admits as much. The point he's making is that it's going to get exponentially worse, until the web is useless (the "tipping point" you mention). What are the "new tools that we already needed" though? I think I'm too pessimistic in my outlook on these things, and would be interested to hear your optimistic future scenarios. Right now, my view is that as that as long as something is profitable, it'll continue. A glimmer of hope is that once the web is completely useless, people will stop using it, and we can rebuild.