Okay, maybe I've got this all wrong.
Is AI a search engine that spits out what it finds on a given subject of already existing information that is currently stored on existing platforms across the internet (therefor it's as good as the information it finds, the old "garbage in, garbage out" problem)
or does it analyze & synthesize variables to form a NEW composite picture that provides unique insight into the requested subject?
For instance, "What effect would finding uranium on Navajo tribal land have on code talkers finding affordable housing in Farmington NM? Is there sufficient water to handle the increase and how much could I get for rents?"
AI should be able to handle those kinds of questions. Otherwise, you're just using search engines that tell you general information, which have been around for decades and can be reached without using Chat GPT or Grok and any other engine.
Or am I wrong? Does AI do that already?
ChatGPT pulls information from the internet.
OpenAI takes 10x more energy output than a Google search anyway.
What's really spook is Anthropic who made Claude AI is now starting to think for itself, build safeguards and has it's own agendas and consciousness.
It other words, Claude has been shown to blackmail, deceive and even have self preservation.
Talk about Skynet in the making.
ChatGPT is just a glorified text chatbot and search engine at this point. It's been proven many times to pull false information from Search Engines. So everyone using it to cheat on a test, are cheating themselves.
ChatGPT pulls information from the internet.
OpenAI takes 10x more energy output than a Google search anyway.
What's really spook is Anthropic who made Claude AI is now starting to think for itself, build safeguards and has it's own agendas and consciousness.
It other words, Claude has been shown to blackmail, deceive and even have self preservation.
Talk about Skynet in the making.
ChatGPT is just a glorified text chatbot and search engine at this point. It's been proven many times to pull false information from Search Engines. So everyone using it to cheat on a test, are cheating themselves.
ChatGPT pulls information from the internet.
OpenAI takes 10x more energy output than a Google search anyway.
What's really spook is Anthropic who made Claude AI is now starting to think for itself, build safeguards and has it's own agendas and consciousness.
It other words, Claude has been shown to blackmail, deceive and even have self preservation.
Talk about Skynet in the making.
ChatGPT is just a glorified text chatbot and search engine at this point. It's been proven many times to pull false information from Search Engines. So everyone using it to cheat on a test, are cheating themselves.
https://www.ibm.com/think/insights/10-ai-dangers-and-risks-a...
Number 10 to me is very interesting
Cool topic! I tend to think of it as "autocomplete on steroids" personally (lol). A massive step beyond google search, but certainly not sentient yet either.
I'll take a stab at explaining my understanding of what goes on under the hood:
1) Large language model chatbots (like ChatGPT, Claude, Gemini, etc) turn the text you type into numbers (vectors).
2)Then they use advanced algorithms to understand the contextual relationships between those numbers.
3) Finally, they synthesize a response based on that understanding.
The responses can appear creative and downright impressive because they are considering a massive amount of information, logical relationships between words and context that the human brain is simply not equipped to handle itself. However, that means there is also the risk for potential "hallucinations" where they make up things that don't work in real life or even make sense. Because of that, it's always smart to apply some good old fashioned common sense to whatever gets spit out at you!
Cool topic! I tend to think of it as "autocomplete on steroids" personally (lol). A massive step beyond google search, but certainly not sentient yet either.
I'll take a stab at explaining my understanding of what goes on under the hood:
1) Large language model chatbots (like ChatGPT, Claude, Gemini, etc) turn the text you type into numbers (vectors).
2)Then they use advanced algorithms to understand the contextual relationships between those numbers.
3) Finally, they synthesize a response based on that understanding.
The responses can appear creative and downright impressive because they are considering a massive amount of information, logical relationships between words and context that the human brain is simply not equipped to handle itself. However, that means there is also the risk for potential "hallucinations" where they make up things that don't work in real life or even make sense. Because of that, it's always smart to apply some good old fashioned common sense to whatever gets spit out at you!