Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror

Comment Boring Books for Bedtime (Score 1) 100

I think this was around before AI
Boring Books Podcast

I was looking for an audio version of a book and find it amusing.
Wouldn't imagine anyway would think The Practice & Science of Drawing, by Harold Speed a suitable candidate.
One of the best books out there, a classic, on how to draw.

Comment Re:Soon (Score 1) 102

If you get a Game, say you have the original DVD, see if it runs on default Wine.
The experience I find is often superior. Because it fires up in an instant.
With Steam, it fires up many things before, Shop, Game Collection, then you having to Log-in.
Which may take a minute or two - or feels that way.

On Steam I have those old Missing Objects games (PopCap) which I gave up playing because can't be bothered waiting for the whole Steam Portal firing up thing. Feels like an eternity for such a light game. Then your "Friends" see you online, keep inviting you to play more serious stuff ie. Left4Dead2 - which can take sometimes 10 minutes to load up "Processing Vulkan Shaders..."
BookWorm, I have it on Wine, takes all but a split second to start.

Lots of old "free" Windows games online.

So Gothic 3, Unreal Tournament 3 (I have the original DVDs) I ran locally.
Quake 1 runs on Linux.
But Quake 2 runs way better on Wine. Excellent games both 1 & 2

The advantage of Steam is that it usually solves the DLL problems with games.
Doing on your own under Wine is often pretty challenging, lots of hunting around for fixes for more complex games. But the instant play after is amazing.

Comment Re:What's next? (Score 1) 52

The reasoning you need to know:
Humans learn Art very differently from AI
Humans witness Nature, Real Life, mostly, in order to be inspired - which is often where they spend most of their time.

Lame Artists may try to copy other Artists - their ideas and styles - and from my observations, these always tend to end up ultra mediocre.
Talented Artists - the ones that are of interest for AI to copy mostly from - acquire their style from research, their own experimentations, inspection into Natural effects; patterns galore, lighting, and so on. And if they study other people's work, it is part of their curricula, and very limited. They know that "too many cooks spoil the broth", so they limit their sources. Copyrighted instructional books (that they paid for) or treatises at the public domain.

AI does *not* need to know: platonic shapes, theory of light and shade, human and animal anatomy (body landmarks), relative proportions (head-height cannon), 1,2,3, n-points perspectives, body types: aging, distribution of fat, material science; pigments, colour theory, composition ... and so on, in order to create more convincingly. Humans do. Can you see the difference?
The more an artist research and learn from various related fields, the better their results, usually. It is a very conscientious discipline, a long strategical game.
About style. This is often one only and the particular mannerism of the artist, evolved from years of practice, of trials and error discovering what works best. Finding a voice. Not by consciously assimilating others. Unless mediocre. People with their own style tend to attract more interest from others than the generic. It becomes their hallmark, their trademark.

The idea that a machine can be excused digesting billions of images, because humans "learn" by merely looking at photos and art from magazines or the web - is complete fallacy. We do not. We have shitty photographic memory and just-in-time analytical processing. We are pretty crap at learning. Keep repeating the same mistakes. Our brains are lazy as fuck, always finding an opportunity to save energy ... Hence the "AI learns Art just like Humans do!" ... a phrase that keeps getting repeated, again and again, mindlessly by idiots.

Slashdot Top Deals

The last person that quit or was fired will be held responsible for everything that goes wrong -- until the next person quits or is fired.

Working...