In some quantum experiments, it was possible to have human thought interfere with random number generators. As if for example imagine at first you have series of millions random numbers forming a “static noise” pattern, once a human performs the thinking action you see a PEAK in the graph (the peak shows irregularity or abnormal results).
So you might ask what this has to do with soul? It has to do with consciousness and not with the mythical-idea of what the soul is. But in terms of consciousness, we talk about the real deal here. Because it has destroyed quantum physics since it was discovered that it can affect experiments.
In some other quantum experiments, it was measured that also a computer, would SEE and affect the experiments (eg: the double slit is affected from computers as well).
So you might ask, how is possible for a lifeless computer (obviously has no consciousness) to act in the same way as an alive human (obviously has consciousness)?
I might say, that perhaps how it works, is that humans oversee computers, where computers oversee the experiment, and thus changing the results.
Or we can just say that those experiments are very inconclusive and we simply don’t know. And as long as there is no clear evidence, I don’t see a reason to talk about abstract concepts such as a soul or consciousness.
I am fine with postulating “consciousness” instead of soul as something that sets us apart from machines of our own creation, though I find it extremely difficult to settle on any further-reaching definition than “sentient self-awareness”. I am also fine with saying that machines don’t seem to have it, as I understand it at this point. Yet. I don’t find it impossible that they might develop something like it, however. I don’t think we’re all that special beyond being highly evolved (for this planet). Complex biological machines. Our consciousness is easily damaged or destroyed altogether by some mechanism or another in our bodies failing.
What little I know of quantum physics seems utterly fascinating, but it is very little, so little that I can’t have an intelligent conversation about it. Sorry!
And all I can think of is how to be reduced to a writer in a command line feeding up some machine and choosing from different options of images for the one I like the most will help me as an artist.
The difference between “Hollywood” and the new generation of the “society saviors” is that instead of having one we have the other earning money. Another difference is that for that other to thrive it decided to stand up over the back of millions of artists around the world and say: now get a job. This is mine.
I would like to remember that patterns of creation as the feeds, words, and sentences “artists” are using there to “create” “art” are also information that’s being gathered. Those patterns are as important, if not more, to the development of the new generation of “thinking machines” as reading a picture or millions of them were before. And guess who is the next to become obsolete?
Now AI and its owners are the portals that will separate “artists” from those who decided not to pay a monthly fee for them.
But as someone said on youtube these days, I’m probably just a craftsman and not an artist, then, irrelevant for that discussion, and all I need to do is to find another way to survive for the next years.
Maybe when I’m sitting on a sidewalk like a veteran asking for some money people will discuss how the “new Hollywood” changed this ill society for the better.
I’m sorry, I didn’t want to be so harsh, but all posts about AI and how amazing they are here are something that turns my guts.
You are right on topic here. In the beginning is all fun and games and at some point things get real…
As Ray Kurtzweil (legendary figure in transhumanism subjects) once said, that if you want to simulate a brain, you can simulate it as a mathematical data structure, without having to go through the trouble of building it in from scratch. Supposedly it would work in the same way… The interesting part is that that the brain consists of billions of neurons, and one neuron is expressed as a sigmoid function connected to other hundreds of sigmoid functions. That’s it! Now you only need enormous computing power!
As for example you would let some virtual critters move around a virtual space and let them do their own thing, all fun and games. The real problem is once they start to compete over virtual resources (evil scientists made them scarce by reducing the integer counters) then you get aggressive behaviors, you get fight for survival and the simulation turns nothing better than the real world.
In this example I just want to point out, that all of these behaviors do not exist in the algorithms, neither there is an if-then-else programming, neither the scientists wanted any virtual creature to get hurt, neither the neural networks are messed up.
Then you have for example “virtual boxers” learning to fight and no more no less for some mysterious reason, they result to fighing in the same way as human boxers do.
But no more no less (for some mysterious reason) these behaviors are starting to get categorized as “emergent phenomena”, ie: behaviors that just happen by no means of intent or grand design but they just are a side effect produced by the overall system.
This is probably the hardest concept to accept for those of us who got decent marks in programming courses in school. First we learn to describe everything as an algorithm, but now there’s this Emergence thing that messes up our perfectly neat world =D
In our own human terms, the entire notion of scientific and academic process, is the process of testing a hypothesis.
If you gradually fail to think and make hypotheses, and not test anything. If you start getting all answers out of a magic black box (the neural network), suddenly you transition into a new era of AI dogmatism.
In terms of speed of result and accuracy the answer that AI gives possibly the correct and most optimal, but the cost is that it comes with more ignorance for the human user.
Back in the 90s (I was about 10) I remember that once automatic text correction became a thing, it was like a godsent technology that would eliminate all possible mistakes and errors. After 20 years or so, by seeing people from my own generation and especially the next younger people, what actually happened is most of them became more illiterate than ever.
Instead of keeping the promise, to correct syntactical errors, it kinda went the other way of making people lazy to write, or irresponsible during writing.
Now just imagine putting AI in every possible place, such as stock market, military systems, etc…
Reminded me of the earliest description of AI emergence I’ve read: "When Mike was installed in Luna, he was pure thinkum, a flexible logic — “High-Optional, Logical, Multi-Evaluating Supervisor, Mark IV, Mod. L” — a HOLMES FOUR. He computed ballistics for pilotless freighters and controlled their catapult. This kept him busy less than one percent of time and Luna Authority never believed in idle hands. They kept hooking hardware into him — decision-action boxes to let him boss other computers, bank on bank of additional memories, more banks of associational neural nets, another tubful of twelve-digit random numbers, a greatly augmented temporary memory. Human brain has around ten-to-the-tenth neurons. By third year Mike had better than one and a half times that number of neuristors. And woke up."
from “The Moon Is a Harsh Mistress” by Robert Heinlein
Sci-fi artists of old (and some ideology architects) seemed to believe in personal responsibility of every human when it comes to their own education. Especially those who were supposed to work with machines were expected to have highly diverse knowledge of sciences, but also other aspects of human life.
Humans were expected to be more… conscious in their thinking and less complacent. To make an effort to be alive.
As a species we were supposed to advance not only the industry but society as well, our way of thinking.
Now while we still continue tech advancement something worrying happens in other areas. We’ve become accepting of bad things, writing them off as “natural”. Food tastes bad? Well, naturally, for that price… Air’s polluted? Nothing to be done here, everyone just GOT to have a personal car these days… Kids are stupid and idle? It’s all those games and Tik-Tok, what can you do…
At this point not just AI but any new tech might have some unfortunate consequences, because we would rather blame technology than people for our character failings.
On a more positive note, if anyone is more interested in AI for Artists than AI vs Artists, Nvidia has a HDRI generating AI that’s pretty cool: http://imaginaire.cc/gaugan360/
Just, uh, be careful with it. Results can be pretty hit or miss. ; D
For the specific niche of “I want a sky and/or clouds with very flat ground that stays out of the way of my actual scene” it’s pretty good though. Gets a little more chaotic when you start trying to use plants or mountains/hills (the latter is due to the editor in the browser more than the AI I think)
Looks like you should be able to make your own “segmentation maps” in a graphics editor and upload them if you download one and color pick appropriately. I haven’t tested it though. (And they’re not kidding when they say you have to wait 30 seconds for it to update, it isn’t instantaneous).
No night scenes though, but if you upload a polyhaven sky hdri jpg as a custom style it works. (If you don’t see your new style, zoom out on the webpage)
If it goes like this, in a few years we would not need texture libraries at all. You would get only a small 20GB “brain” and let generate unlimited amount of possible textures.
We have procedural tools for texture generation, like substance designer. It’s not AI, and actually requires a good deal of technical skill to use effectively, but is comparable for how it has affected how many, many textures are produced and processed - including for texture libraries. Instead of replacing them completely, substance designer and other procedural texture programs and techniques exist alongside and in combination with photo and scan based methods, and even texture painting from scratch.
I think the best AI tools will always be the ones artists have the ability to direct themselves, which means it will still likely take skill to use effectively. Especially for something like textures where you’d want AI to increase your control over your texture creation process more iteratively and efficiently. Unlimited texture generation isn’t particularly useful if the results are too random and unpredictable if you’re looking for something specific.
Plus I suspect an AI based texture generation tool would likely be reliant on new textures being input into it too as part of the process, since it seems unlikely one could come up with a training data set expansive enough to cover all possible texturing needs - people can get quite particular about that stuff, sometimes even traveling to get scans of the exact right kind of rocks, dirt, bark, whatever for a specific geographic area.
Bumping this thread, because it looks like AI-generated art made in a few seconds could pose a serious thread to artists who spent days of their life pouring their heart and soul into a piece (at least when it comes to contests).
In a way, this cheapens the concept of contests in general, as no one will want to actually create something manually if this becomes the norm (shred those canvases and trash those brushes, you will no longer have a chance, ever).
Art contests are hot garbage anyway though, and deserve devaluing, so I am not going to get het up about that aspect. But it’s a fine test case regarding “but is it art” – apparently the Colorado State Fair judges thought so. It’ll stir the public debate, and that’ll be interesting and possibly positive in the long run. Possibly. Being aware of possibilities is a good thing, but man, people so often overreact in all the wrong directions. Still, better to have the debate than to be blind-sided.
Besides, they have different classifications for their hokey contests already. This piece won in the Digital Arts/Digitally Manipulated Photography category, so people who paint on canvas with brushes had nothing to fear from it. Digital art in general is already assisted by a multitude of algorithms, why would AI-assistance be different? Where do you draw the line? People will try, of course, if this gets enough of them upset enough to want to protect human artists.
And as regards “AI-generated art made in a few seconds” – no. The artist who submitted the artwork said he worked on it for weeks, exploring his prompt, generating hundreds of images, curating and fine-tuning them. Human effort was expended here, Midjourney didn’t wake up one fine morning and created this all on its own in a few seconds. At least not at this point.
And don’t forget it happened in Colorado- which, don’t get me wrong, is a nice place, I throughly enjoy Steamboat Springs, but it’s also bordering Wyoming, Arizona, Oklahoma, and Utah. It’s pretty close to Idaho as well. I refer to that part of the country as the “ugh not you guys again” belt
Disclaimer: I live in Oklahoma and moved here from Idaho so I’m allowed to make fun of these places
But as they say a picture/3d model is worth 1000 words so unless you want to type in a novel as the “prompt” to the AI then you may as well learn to draw/model yourself to get exacly what you want.
AI is just the latest tool for artists. It can’t replace them. When I started as an artist, there were no good printers, in fact there were no good computer graphics of any kind. Most graphics were done by hand using rulers and pens and brushes. People don’t hire me to do hand lettering or hand painted signs anymore, but that’s ok, because I adapted to the new technology as it appeared. Don’t fear technology. USE it.
I wonder if the judges were people who understand that digital artwork is still work to create, or if they place so little value in the time and skill it takes to create digital artwork, that they don’t see the difference. Some people seem to think the digital stuff comes out of thin air normally.
on all the AI artworks, I really get the feeling that it is all about approximation. Because the closer I look the more pieced together it seems. It’s like it is lacking connecting details.
Either way you look at it A.I. is still a Journey away from creating 3D art from one’s Mind.
