This is exactly why I feel nervous asking questions online. I feel like a lot of the time the answer is so obvious that a bot could answer it with very little context and then I’ll look silly.

Sometimes it’s just like monkeys on typewriters

That is a Savage response. As for being uncurated, who knows…

the AI is smart enough to write a decent response, but will it ever truly be smart enough to know that arguing online is like pissing into an ocean of piss?

GPT4 reply:

I think your comment highlights a significant and often overlooked part of the AI creation process. In many cases, the output we see from AI is indeed the result of numerous attempts and substantial human curation. The “trial and error” part of the process is not typically highlighted, but it’s a critical part of AI development.

It’s worth noting that the use of AI in creative applications like this one doesn’t diminish the necessity of human skill and judgment. In fact, it emphasizes the complementary role that humans and AI can play together. The AI can generate a vast amount of diverse and novel content, and the human can act as a filter, curator, and finetuner to produce the final result.

So while headlines like “I made an AI do…” might create an impression of complete autonomy, the reality, as you pointed out, involves a significant amount of human input. It’s the synergy of AI capabilities and human discernment that makes such projects successful.

That being said, with advancements in AI, we are seeing improvements in AI’s capability to generate more consistent high-quality output, but it still needs human interaction to guide it towards a goal that aligns with human values and aesthetics. That’s why it’s important for AI developers and researchers to be transparent about the process and the level of human involvement. This helps set realistic expectations about what AI can and cannot do, and emphasizes the continuing importance of human creativity and decision-making in the AI-powered future.