LWPs will be used indiscriminately and for evil

THE AUGUSTE LAURENT SOCIETY @JorgeJi83882694  The AI will put us ahead though; do the work of naming, sorting, clarifying. But Rising tide will not lift all boats. Simply dont underestimate their ability to make up work and graft and graft. People are more inflamed
Replying to @JorgeJi83882694

In every case you see ( LLM wrapper program (LWP) does something good), it is because the human has guided the conversation. When you are not familiar with the subject, and the LWPs offer untraceable statements and declarations from those statistical memories, much of the time the human has no way to check.
So errors can easily infiltrate the Internet, and human knowledge. It is worse then a deliberate virus, because it is impossible to trace. The LWP makers will churn out lies and falsehoods faster than any diligent person or group. The politicians only care about winning a temporary prize or 15 second of fame. And the LWP makers are giving away tools, to more people – perhaps with less scruples will churn out still more lies and falsehoods.
It will be nearly impossible to detect, to reverse or root out. The patches and fixes will be too late. The LWPs are built on the wrong principles. It is not impossible to correct, but with many billions of dollars and a few hotshot leaders pushing so hard, if will be “on the backs of the many and weak” again.
If the groups would get “all languages”, “all knowledge in lossless form” and “use global open tokens (words and phrases) as translation ids”, it might change even now. No one is beating down my door, and it is hard to explain decades of quiet work, when what sells is hype, violence and conflict. The groups will have alternate ideas, then posture to sell themselves to the highest bidder.
Filed under ( LWPs will be used indiscriminately and for evil)
Richard Collins, The Internet Foundation

When LWPs (LLM wrapper programs) are classifying, sorting and clarifying they use inefficient and expensive generative methods when there are much more precise and low cost machine learning and lossless database methods based on probability, statistics from STEMC and other professions – to work from.

A tiny few LWP makers are trying to grab a global market with “artificial” methods and NO guarantees or oversight. They use “we have to use giant computers so it is expensive” and “we are too intelligent and smart to be bothered with curating and verifying and tracing the input data”. So they are not “giving anything back” to the humans they steal the knowledge from. A thief is still a thief, even if they are not seen or caught

 
Richard Collins, The Internet Foundation
Richard K Collins

About: Richard K Collins

The Internet Foundation Internet policies, global issues, global open lossless data, global open collaboration


Leave a Reply

Your email address will not be published. Required fields are marked *