The custodian for all human knowledge

I was watching Bill Gates' new documentary - What Next on Netflix today. The first episode, perhaps unsurprisingly was about AI. It was on a whole, a well-rounded, hopeful take on AI.

There were two sections which jumped out at me.

1. It was only recently that the amount of content produced by us was enough so that the LLMs can guess better.
2. It is basically trained on the web.

It is trained on the web. All the stuff that is on the web.

In an altruistic sense, I am OK with using the world's knowledge for the benefit of humanity. For training something that can help us tackle a lot of things, which we frankly can not.

However, the world's knowledge should belong to the world. Not to corporations chasing profit. OpenAI has said as part of its recent funding round that they will be becoming a profit-chasing/making company within two years.

There could be an organization, perhaps the United Nations, which becomes the custodian for all human knowledge. Books, articles, images, everything.

Now if a corporation uses this data, to make profit, then there's no altruism in that. Then there should be a way for the creators to make money as well. I don't know how commerce on the web should work. Subscription fatigue is real. Maybe you pay a flat fee to access everything on the web. And that money gets distributed to everyone.

The web is a mess right now. Search is shit. There are so many content farms filled with AI generated slop. Google traffic is tanking, or has tanked for a lot of publications.

I have a list of websites/blogs that I have open all the time in my browser. Each day I go and refresh and read whatever new is published. I understand that is not how the rest of the world works.

Distribution is the problem.

Fediverse maybe a solution. Or a part of the solution. Wordpress has enabled fediverse integration. Ghost is working on it as well. So, maybe, this distributed future may provide a solution for the distribution problem.