The majority of the internet is porn.

Again, I’ll separate entertainment from informational, since entertainment can be garbage, and still be consumed.

Bad information doesn’t help anyone.

it’s not like LLMs you can chat with are completely useless.

The problem is, you wouldn’t know unless you know.

With a legitimate website that has human writers, editors, and fact-checkers, they can at least have creditability and a reputation to uphold.

Far too many randomly generated websites have a lot of information, but without any guardrails. If you know enough about a topic, you’ll realise that the information on these AI sites are pretty much useless. That is, you couldn’t use them as a source because enough of the info is bad/incorrect/incoherent, that it’s like asking a toddler who may or may not give you a valid question.

I’ve contacted a manufacturer of bike stuff, and their support is given by AI. While the answers you get sound like they could be right, it’s like getting an answer from someone who heard something about something from a friend. When you actually ask for a human, the answer is often different (and correct).

There is no accountability, or credibility, or responsibility, or integrity with AI. It has no reputation to lose if the information it provides is bad or not.

I know that AI isn’t going away. I’d personally be OK with some human verification system for websites, and would be more than willing to use a filtered version of the internet that blocks AI generated content. Call it curated or whitelisted, but I want my information to come from a human being.