So this is kind of interesting, it doesn’t look as terrible as a lot of AI animations I’ve seen. There are still some weird morphing elements in some of the shots, but a lot of it looks OK, at least in terms of consistency. However the limitations are obvious. The characters barely move at all, and it seems to be mostly head and shoulder shots rather than showing the full body. Its actually kind of insulting to Pixar that this guy compares himself to them. These characters are completely lifeless compared to anything you would see in Pixar movies.
1 Like
Actual creativity on display in DCC apps. might end up becoming just as much a lost art as cursive writing, map reading, and mental math calculations. AI provides a lot of new ways to outsource human thinking and problem solving so the brain can rot even farther (for instance, if you think the so-called phone zombies are bad just wait).
If all of the technology goes away for some reason (blackouts, CME, ect…), you will be at an advantage if you remain among those who actually know how to live life without needing Google or ChatGPT on hand to give instructions. This might actually be a self-fulfilling prophecy as humanity loses its ability to keep the march of technology intact (because the end of the need to think will have reverted society to a far more primal state, better have a plan to escape the cities and head to the nearest Amish village when that happens because they will be among the few civilized places left).
Looks like the Japanese government started using GPT for administrative tasks. Soon or later for what else?
https://the-decoder.com/microsoft-brings-chatgpt-and-metas-llama-to-japanese-government/
Typically, this would not ring any bells. But the problem is that I was listening to a podcast of Peter Diamandis the other day, that said that there is a prospect that in the future governments would be fully automated using A.I.
Kinda reminds of that game “Beneath A Steel Sky” (warning very dark game, do not play it).
It seems they are using the capability of large language models in a way that is know to help (at least many humans who tried to use the tool) to work faster.
Moving those systems from assistant mode to fully automated would require massive improvements. It might work in some niches in the case they regularly retrained the models with the data that is being produced.
In my opinion, “it could be that this and that happend” is not an argument. Pretty much every tool we have on computers has seen massive improvement over the years to help people solve their tasks. Now the language assistants are going beyond typos and grammatic hints and they are able to make suggestions.
So is that artificial artificial intelligence then?
AAAAAI
1 Like
Yeah, I remember Kaedim from when they made a splash last year. I thought they were using human artists even before they came out with their blather about “quality control”, it was pretty obvious from comparing wireframes that different models had been created by different artists rather than by an AI.
Of course when there is this much hype, and so much money to be made out of the gate, there’ll be people lying.
2 Likes
There were rumors about that from the beginning, but now it seems confirmed. I guess a lot of scammers are riding the AI popularity wave…
2 Likes
They were featured on the Forbes 30 under 30. That’s all you needed to know.
For those who aren’t aware a shockingly high percentage of people featured on the Forbes 30 under 30 End up being scammers.
2 Likes
A top newspaper publisher in the US has just deactivated its AI for writing articles on high-school sports.
USA Today’s publisher had to update all of the sports posts its AI reporter botched (engadget.com)
This is the publisher for USA Today, the paper that you find at the door of every hotel room Monday through Friday. They thought the boring task of tabulating scores would be something they could outsource, only to overlook the nuances that separate the reporting of this group from College and Pro sports.
1 Like
Whenever I read about those sorts of cases, I wonder who in the first place would come up with the idea to let large language models write the articles and then not properly test whether it actually works. The only answer I could find so far is money driven idiots.
3 Likes
Better to be prepared for more of that …
This looks like the RECAPTCHA of the 3D-MODEL-GENERATION.
Those are the kind of people who take every opportunity to make money, no matter what. With all those AI tools, they found some new toys. Personally, I don’t see a need to be prepared for the stuff they come up with, but it is more important to be aware and avoid those people.
(post deleted by author)
I can confirm some of these “rumours”, i know someone that is selling AI books on amazon already
we can already see some instagram pages, full of fake AI pictures
a crazy world we’ll live in…
It’s everywhere
500 000 likes and people don’t see or don’t seem to care that it’s fake
*sorry for the double post, folks
I recently started reconnecting with various creative networks around the web, and I quickly became aware of how big a deal AI has become. Not the whole “AI makes stuff, oh no” debate, which I obviously know about, but the issue of AI training companies ‘scraping’ images from the web, i.e. using automated programs to copy images, often without consent, that they then train their AI on. This means that your stuff could end up being the basis for some AI image generator, without you ever being asked for permission.
It is a pretty big debate, I now know, and a lot of artists are refraining from posting their work, terrified of being essentially ripped off by machines. How do people feel in here about this issue, and what should be done about it?
Most people are against such practices from what I’ve seen. I am not completely anti-AI, but I am against the scraping aspect. The current AI would be useless without the human output, so I think the data collection part of AI needs to be heavily scrutinized.
I feel that the market will regulate this by itself. If AI truly replaces people without anyone noticing or caring then that will be it. If people get tired of the “AI look” or demand the hand-made component, artists will remain in business.
Within our lifetimes we’ll probably see no more and no less than some in-between compromise. The current events show that it’s still too difficult to integrate AI tools with on-premises workflows in a manner that brings forth the proclaimed industry revolution.
Perhaps more relevant, the ego-driven revolt of artists against AI testifies to how much artists have yet to accomplish in the art of living life… 