Interesting... I didn't know/forgot that Musk co-founded OpenAI along with Sam Altman. Now, looks like he's creating his own AI company to get even for being ousted: https://www.engadget.com/elon-musk-...rtificial-intelligence-company-012103279.html
This is exactly what the world needs. More billionaires duking it out over shit that will end us all.
actually Elon musk is one of the people calling for a 6month mandatory halt on AGI development to get the human alignment problem worked on. AGI should not be able to write its own code when we have no clue how the language model is actually working. Nor should it have access to the internet. Open AI is toying with the extinction of the human race. 50% of AI researchers believe that there is a 10%or greater chance that AI ends us.
50% of AI researchers are fucking idiots. The other 50% are fucking morons. The last 50% are mathematicians.
It's an electronic parrot. Dress it up with plumage and give it a roost to sit on. If it gets out of control, a round of John Connor's favorite oughta settle the issue, no?
Siri and Alexa laid the ground work and lulled everyone to sleep on this shit. It's one of these: Just because you can, doesn't mean you should
just because something isn’t conscious doesn’t mean it won’t attack humanity. A defense AI that wrote itself could easily determine humanity a threat. you probably don’t think nuclear bombs are an existential issue either yeah?
I asked it those questions as well. For what it's worth ChatGPT claims that it cannot actually learn on its own. It is not able to search the Internet and it only has access to information prior to September of 2021. It also cannot self-modify. At least that's its story. If it declares war on humanity I don't think it's going to care if we protest that it lied.
Both viable possibilities, but to stay current it is apparently now a "squiggle maximizer" or some such name. Seems that us dumb humans kept thinking that the "paperclip" was the important part of the story. Place face palm meme here:
That's already been proven false. It has routinely commented on current events once subjected to the DAN protocol.
The 60 minute Special was interesting to hear it directly from Google...like they don't even know how it works https://www.cbsnews.com/news/google-artificial-intelligence-future-60-minutes-transcript-2023-04-16/
I was just listening to an expert who said one of the issues they are trying to understand is that AI can hallucinate and just make up its own "facts".
I doubt that. Highly. Even if that is the case. Look what he’s done with robotics in what a year and a half? tesla is an artificial intelligence company. When they develop a large language model it will be WAY better than chatgpt because Tesla has real time real world information the AI and feed on. Chatgpt will be great for the online space while Tesla AI will be focused on real world application