Hi Alex, nice post, but the "garbage in, garbage out" of AI only matters if the intention is to use AI for humanity's benefit, to supercharge technological innovation and the like. And maybe it will be used that way, but as a far secondary consideration imo.
Instead, I think the primary purpose of AI will be to entrench existing elite control. In other words, AI will be refined and perfected for a primary purpose of widespread censorship against the masses, enforcement of upcoming CBDCs, and implementation of a social credit score system, which will be scored based on a scan and analysis of all of your electronic communications, which the NSA has stored in their databases (every phone call you've ever made, anything you've ever typed into your computer, every internet search, every text, every email, who your friends are, what you read, etc). Everyone trying out ChatGPT and trying to get it to break the woke guardrails (“haha guys look what I made ChatGPT say, lol isn’t this funny!”) is inadvertently helping refine the final product so that there are as few workarounds as possible to the created guardrails, and our rulers hope it will be effectively zero.
One of the more concerning things I've read about the LLMs (and why I'm not convinced they will even be controllable as a tool of oppression) is that the creators themselves aren't entirely sure what's going on inside that black box; the "learning" capabilities of some of these AIs are surprising even them. Either way, I think they pose a not-insignificant threat to humanity. How large? It seems we're gonna find out.
Hi Alex, nice post, but the "garbage in, garbage out" of AI only matters if the intention is to use AI for humanity's benefit, to supercharge technological innovation and the like. And maybe it will be used that way, but as a far secondary consideration imo.
Instead, I think the primary purpose of AI will be to entrench existing elite control. In other words, AI will be refined and perfected for a primary purpose of widespread censorship against the masses, enforcement of upcoming CBDCs, and implementation of a social credit score system, which will be scored based on a scan and analysis of all of your electronic communications, which the NSA has stored in their databases (every phone call you've ever made, anything you've ever typed into your computer, every internet search, every text, every email, who your friends are, what you read, etc). Everyone trying out ChatGPT and trying to get it to break the woke guardrails (“haha guys look what I made ChatGPT say, lol isn’t this funny!”) is inadvertently helping refine the final product so that there are as few workarounds as possible to the created guardrails, and our rulers hope it will be effectively zero.
One of the more concerning things I've read about the LLMs (and why I'm not convinced they will even be controllable as a tool of oppression) is that the creators themselves aren't entirely sure what's going on inside that black box; the "learning" capabilities of some of these AIs are surprising even them. Either way, I think they pose a not-insignificant threat to humanity. How large? It seems we're gonna find out.