Welcome aboard, dear readers! This is your pilot speaking, and I'd like to welcome you all to today's flight, destination: automated AI writing. In our last post, we explored the AI writing assistant, and today we're taking it a step further by handing over the controls to our AI writer co-pilot, testing the limits of automated creativity at 30,000 feet.
I love being able to create an outline from a mess of ideas, but it's important to remember that these are tools. If the process is "turn the keys over to the AI", you're going to get utterly generic output every time. That said, this is a really interesting experiment!
I agree; I prefer the collaborative process - where every step can be made better through the back and forth. But I was surprised at how even a fully automated process can be made better through enough guidance and examples.
It's safer to assume that AI can and will reach 'non-generic' outputs. It can already mimic anything it was trained on, like the styles of famous writers and artists. Say you wanted to achieve 'high-quality X' for some X. With enough instructions and examples of what 'great X' looks like, I don't see any reason why it couldn't produce an excellent output for any X.
This is already within reach with every new model, and you can fine-tune open-source models on any specific task. And with longer context lengths, all the 'training' you would need is to give detailed enough instructions and examples in your prompt.
I love being able to create an outline from a mess of ideas, but it's important to remember that these are tools. If the process is "turn the keys over to the AI", you're going to get utterly generic output every time. That said, this is a really interesting experiment!
I agree; I prefer the collaborative process - where every step can be made better through the back and forth. But I was surprised at how even a fully automated process can be made better through enough guidance and examples.
It's safer to assume that AI can and will reach 'non-generic' outputs. It can already mimic anything it was trained on, like the styles of famous writers and artists. Say you wanted to achieve 'high-quality X' for some X. With enough instructions and examples of what 'great X' looks like, I don't see any reason why it couldn't produce an excellent output for any X.
This is already within reach with every new model, and you can fine-tune open-source models on any specific task. And with longer context lengths, all the 'training' you would need is to give detailed enough instructions and examples in your prompt.
It's going to be wild when an AI can compose a novel better than 99% of published authors!