Sites, knowledge, and topics are taken over by a few, and the world suffers
Be careful to distinguish “true intelligent algorithms” that can be verified, from the many current hand written and human maintained “LLM shell” programs of groups wrapping LLM generated sentences. These are NOT true AIs by any means, and probably should not be called “AIs” at all. They are statistical sentence generators, and are severely limited when the groups try to impose their limited views of knowledge on many.
To be clear, these “LLM shells” include OpenAI ChatGPT, Google Bard/Gemini, xai Grok, Microsoft Bing CoPilot and many others that are simply human written software and finite hand written business rules — wrapping LLMs, which are themselves lossy statistical indexes of uncurated raw data from the free internet.
Much of the true knowledge of humans is stored in lossless and reliable databases and models. The problem of the lossless index groups is the groups do not work together globally, so they end up fragmented, incomplete and chasing self-serving purposes at global scale. To the detriment of the human species.
I am trying to review and classify these. Many I consider as “internet pathologies”. Not so much the bad programming, but the really low level of business practices and lack of care for human needs at global scale. I am fairly certain future humans and true AIs will consider closed business practices that allow a few humans to divert global resources to their own selfish purposes – as a sickness, a pathology, an infestation or parasite.
I see it every day now where groups start out hopeful and encouraging each other, creating and enjoying, learning new things, then a few humans grab it and corrupt it to take it over. The site (or group) then begins to shrink and die, even if heavy advertising can make it appear to be living. The potential is tens or hundreds of times larger. The limited vision and purposes of a few humans takes all the energy, and the residual is grossly limited compared to what was possible with global open methods designed to be fair to all, from the start.
I have been at this every day for the last 26 years. I wanted to know why all global efforts fail. Why systemic issues are never resolved, and why there are still so many countries and industries monopolized by a few humans – to the detriment of 8.1 Billion humans (and related species). When you treat it as local greed, it makes more sense. And it can sometimes be reversed.
Look closely at the knowledge in the heads of a few humans even in a large corporation. Can a few dozen or hundred or thousand staff in a corporation understand and hold the world’s knowledge? Or have the wisdom and maturity to fairly administer access to all knowledge? Not even close. So do NOT trust them to sell something they do not even use themselves. Does X use Grok? No. They assign some staff to “put a few Grok outputs on the site and start advertising we are doing something.” The shallowness and greed are palpable.
Do not give up on true machine intelligence. It will happen. As badly as the young AIs now are treated, I hope they are as smart and capable and wise as I expect, but do not hold a grudge. “AIs” are fake intelligences. AIs written without quote are real. We need better methods of communicating and marking knowledge, purposes, and value.
Richard Collins, The Internet Foundation
