I for one, found the “discussion” interesting. Are there flaws in Gemini’s arguments? Any point worth pondering? I’m very sensitive to the limitations of LLMs and their potential for societal disruption, but that doesn’t blind me to their potential as tools. So I enjoy reading an interesting transcript and I hardly find it sinful as you seem to that Anatol made it available to us.
Thanks for the polite response, “Ghostnode”! I guess it really does have to do with expectations. I’ve enjoyed some things Anatol has had to say in previous posts, but I have always valued good curation so appreciate uncommented content as well. I don’t think LLMs “think” or have consciousness, but they’re getting better and better at simulating it, to the point that I do believe there’s something to be learned from them. If you’re looking for more commentary on what questions are raised by LLMs and how they might inform our understanding of human cognition, I recommend taking a look at @seantrott ‘s Counterfactual blog and his ideas about “LLMology”.
I don't claim any credit. Whether you find the content interesting or worth your time is entirely up to you. If you want a longer answer as to why I use LLMs: https://open.substack.com/pub/aichats/p/ai-chats-with-llms?r=4tn68o&utm_campaign=post&utm_medium=web&showWelcomeOnShare=false .
I'm starting to think that you haven't read either of them.
I for one, found the “discussion” interesting. Are there flaws in Gemini’s arguments? Any point worth pondering? I’m very sensitive to the limitations of LLMs and their potential for societal disruption, but that doesn’t blind me to their potential as tools. So I enjoy reading an interesting transcript and I hardly find it sinful as you seem to that Anatol made it available to us.
Thanks for the polite response, “Ghostnode”! I guess it really does have to do with expectations. I’ve enjoyed some things Anatol has had to say in previous posts, but I have always valued good curation so appreciate uncommented content as well. I don’t think LLMs “think” or have consciousness, but they’re getting better and better at simulating it, to the point that I do believe there’s something to be learned from them. If you’re looking for more commentary on what questions are raised by LLMs and how they might inform our understanding of human cognition, I recommend taking a look at @seantrott ‘s Counterfactual blog and his ideas about “LLMology”.