Chatgpt weighs in on an age old question..

Well, at the moment, LLMs combined with natural language programming are changing things, and ChatGPT is the way the public has become aware of them. Certainly not a fad. But at the same time, this is the fourth or fifth time in AI history that people have announced THE FUTURE.

We’re still years and years away from AGI, from fully autonomous vehicles, etc. AI sentience might never happen. Don’t let the magic 8-ball fool you.

AshyKnuckles

ChatGPT is starting to be the new annoying fad
It's the FUTURE
Well, at the moment, LLMs combined with natural language programming are changing things, and ChatGPT is the way the public has become aware of them. Certainly not a fad. But at the same time, this is the fourth or fifth time in AI history that people have announced THE FUTURE.

We’re still years and years away from AGI, from fully autonomous vehicles, etc. AI sentience might never happen. Don’t let the magic 8-ball fool you.

Well I get lonely sometimes so this AI interaction is groundbreaking for me lol 🤭

Artoria77

ChatGPT is starting to be the new annoying fad
It's the FUTURE

Casual reminder that this is real life not some sci-fi movie 😭💀

BuggyDClown

Lol the new meta is not the best meta according to AI 🤣🤣

Seeing Void in that list makes my day 👍👍👍👍

HSS75

MrSakuragi

DNA3000

ChatGPT is starting to be the new annoying fad
It's the FUTURE
Well, at the moment, LLMs combined with natural language programming are changing things, and ChatGPT is the way the public has become aware of them. Certainly not a fad. But at the same time, this is the fourth or fifth time in AI history that people have announced THE FUTURE.

We’re still years and years away from AGI, from fully autonomous vehicles, etc. AI sentience might never happen. Don’t let the magic 8-ball fool you.

I've lived through the expert system era, the early neural network era, and the semantic knowledge network era of AI. I think the current state of the art of LLMs and deep multimodal transformer networks in general feel different to me. In every past era of AI, the AI technology was sold in terms of its potential. LLMs are being sold in terms of what they can already do. Which is not to say they are perfect, but I don't think this is going to be one of those times when we look back at the Palm Pilot and think this was cute. I think we're going to look back in the same way we now look back at the first iPhone.

When you're living in the moment, it is not easy to tell the difference between the Palm Pilot and the iPhone. But I don't think ChatGPT is a Palm Pilot.

LLMs and transformer networks are literal game changers, in that they have literally changed the game. The goal of strong AI has always been to develop systems that were equal to or superior than human intelligence. ChatGPT has shown that this is neither a necessary nor sufficient requirement of strong AI, because humans are incredibly bad at judging whether an AI is equal to or superior than human intelligence. ChatGPT proves the fundamental flaw of the Turing test is there are no officials qualified to judge the contest. Deep down, that's what has many AI research observers nervous. The iPhone changed the game by inventing a new one everyone decided to play. Even if you don't own an iPhone you live in a post iPhone world where everyone plays the iPhone game. Apple's greatest rivals try to beat Apple by playing the iPhone game better than Apple, and the world they collectively create to win this game is an unrecognizable world. When you look at the first iPhone, you don't see Instagram or Tik Tok, or IRL Twitch streaming, or swipe right dating. Those are the things that ultimately change lives.

Few if any of us ever used an expert system. Expert systems can change our lives, but only in invisible ways. LLMs are likely going to change things in visible ways that people will interact with and build upon. That's when a technology has the chance to become globally game changing. The automobile did that, the internet did that, the smartphone did that. LLMs are the first AI technology that might do that. Either way, we'll probably know in just a few years one way or the other. I'm betting they will make the cut.