Recent searches
Search options
Oregon Coast Trail @killyourfm@layer8.space*Groan*
And so it begins in full force.
Today, both Mailchimp and Hootsuite (very popular tools which I happen to use for @thunderbird newsletters and social media) invited me to use #AI to compose Facebook posts, tweets, and emails, respectively.
I won't be using AI tools for writing. But you should know that it's starting to get exponentially more difficult knowing if the content you're absorbing is from a human or from something like #ChatGPT.
@killyourfm @thunderbird Personally, I think that is bad since it will only increase the flood of shitty "SEO optimized" bullshit spam & -posts...
@kkarhan *Nods*
It's about to get really ugly out there.
@killyourfm it also really makes removes any value from newsletters and mailings they had - I'm not a big fan of either but that doesn't mean they have no merits...
@killyourfm My favorite content is from a human who used #AI :)
@ned I love certain AI-driven applications! Like Midjourney for example. I'm just saying that AI could never capture my unique writing "voice" and style.
Or at least, I hope it can't..
@thunderbird Sure, but it can make something somewhat close to your style. It will output much more/faster than you can, but the quality will be quite different. I edit a lot more than I write, so LLMs are a great tool for me. They're kinda useless for pure writers.
@ned I could definitely see using LLMs for idea generation and research on my end.
@thunderbird if you're a fiction writer, instruction-following LLMs can help you prototype possible dialogs. This is what character.ai is about now. Perhaps you can set up a character based on your writings? #ChatGPT is more like a postal office employee. Too formal for non-technical writing IMO.
...and let's wait until we're microchipped or something similar - and this tech is assimilated to that chip "to aid" human interaction 
well...
@killyourfm @thunderbird you’ll probably know when you ask follow up questions in person and it’s clear they’ve stopped proof reading what ChatGPT has been generating for them. 
@killyourfm @thunderbird Usually sentences from ChatGPT are repetitive and the structure is wrong (even if not detected by zerogpt).
@killyourfm @thunderbird Today we got the first mail very obviously written by AI. It didn't fit the tone of the previous email conversation at all. 
@Halefa @thunderbird I don't know the details, but it did it make you feel disconnected to the brand or the organization sending it?
@killyourfm @thunderbird I use ChatGPT to reply to LinkedIn recruiters.
@killyourfm @thunderbird it's even a bit irresponsible as the copyright status of AI output is unclear. What if months worth of your social media posts turn out to be plagiarism?