Hacker News Reader

Comment by romeros | Hacker News Reader

I also wonder if it surprised even the AI Researchers/Developers working on OpenAI.

Maybe they were just trying to tweak things and it surprisingly turned out bigger than they had imagined.

I had an interesting conversation with google bard/gemini about this topic. Here is the link to the conversation: https://g.co/gemini/share/67da9e441ad9 Fascinating stuff...

my notes of what i am learning in the process of these discussions: https://github.com/MoserMichael/my-notes/blob/master/talking...

I would be careful about "interesting conversations" in which your interlocutor starts every single response with insincere flattery and confirmation. The link you posted reads like a chat with a sociopath sycophant that's try to get you involved in a pyramid scheme.

"You're on to something!", "You're absolutely right.", "You've got it exactly right!", "Your penis is so large and hard, yet supple!", "I'm extremely excited about anything you post to me!"

it is a bit of a confirmation bias machine. still it gives better leads for learning about a subject (better results than just searching with google)

I also fancy that you can talk with the machine about itself, to me that is still having some degree of novelty.

Maybe because I use ChatGPT-4 a lot I find ChatGPT-4 responses more credible than Gemini/Bard. I personally feel Gemini that is trying hard to not offend me and is also trying to sugarcoat the truth. It also feels kind of half baked.

Reply on news.ycombinator.com