
The Screen Actors Guild strike is now in full throttle, with artificial intelligence a key component of the walk-out. (SAG rejected a proposal by Hollywood studios to create and own AI versions of actors in perpetuity without compensation.) Inevitably, that's caused some AI evangelists to wonder: Do we even need human actors any more at all?
"Doesn’t this just create demand for Hollywood studios to create Synthetic Actors?" As my colleague Jeremiah Owyang put it recently. "Disney+ Netflix: seems like over 50% of content is CGI or digitally altered."
There's already quite a lot of digital-only actors in movie and TV, especially in epic crowd scenes. (And for quite some time: WETA Digital innovated this for the battles in the original Lord of the Rings trilogy over two decades ago.) So the real question is whether we should expect studios to clamor for digital-only actors in bit parts -- or even in starring roles.
There I'm high skeptical, not just on a "Do we really want all human culture synthesized?" moral and philosophical level, but on basic feasibility:
Wonderful: ChatGPT Apparently Improves My Blog's Comment Spam!
One super-fun Monday activity i have is going through this blog's comment filter to confirm the nuking of apparent comment Spam. For the longest time (like the last 15 years), it's been quick and easy to spot the Spam comment, which is totally off-topic promotions with a link to a commerce/scam sites. In a recent innovation in Spam comments, a bot apparently mixes up the Spam text with garbled excerpts from the individual post, so it takes a bit more time to identify it as Spam.
Now, evidently thanks to ChatGPT and other Large Language Models, blog comment Spam has taken a great leap in quality!
Pictured at right are a couple recent Spam comments, similar to several others that have been popping up on New World Notes recently. Unlike the Spam of the past, the comments are actually on topic to the specific post, and grammatically correct. (If totally bland.)
It's only when you check the commenter's URL do you realize it's a Spam bot promoting gift cards and obscure software or whatever. (Don't bother looking for the actual comments, they've already been nuked.) Ironically, many of these Spam comments generated by AI are attached to posts about AI.
So that's wonderful! It's great getting to waste an extra 5 minutes a week on comment moderation, so thanks for that, OpenAI.
There are some broader implications here:
Continue reading "Wonderful: ChatGPT Apparently Improves My Blog's Comment Spam!" »
Posted on Monday, July 24, 2023 at 02:34 PM in AI, Comment of the Week | Permalink | Comments (1)
|
|