man_in_space

Just for fun, I threw your username phrase into it. First two are base, send two were called with --niji.

deleted by creator

See what I mean about the AI not taking criticism?

Jeff getting shot in Uncharted 2. He at one point gets wounded and Drake (the player) has to do the shoulder-carry-walk thing with him in the middle of an urban attack. He’s in a bad state and isn’t the fastest, but he’s whatshername’s coworker and you’re his only hope out of there.

And then right when you get to your destination, there’s Lazarevic, and he shoots Jeff dead.

Yes. For the posted image specifically, I requested a Norman Rockwell-style painting.

With Midjourney at least, it gives you four options and the ability to alter (“vary”) those options. If none suit your taste, you can rerun or change the prompt. In addition to specifying things in the prompt itself—there’s a wide latitude, left to its own devices it defaults to a kind of digital art style but you can override this with hard-coded options and/or specifying genres and media; I frequently use “oil painting on wood”, for instance, and you can specify styles (cubism, op art, abstract art, futurism, newsreel, photorealistic), color palette (dark, light, vaporwave, monochrome), media (giclée, daguerreotype, airbrush, tempera, canvas, wood, tintype, charcoal, ukiyo-e), and artists (da Vinci, Bosch, Escher [Escher produces fascinating results more often than not], Tanguy, Mead, Beksiński)—there’s a few arguments you can apply to the query. The one I remember offhand is, I think it’s called, “niji”, which specifically forces an anime-esque aesthetic.

DALL•E is similar, even showing you four options you can vary—but it’s much less robust in what it can do.

The tradeoff is, and this is why AI won’t be some silver bullet, the AI can’t take criticism. It can only handle so much information at once, and you’re kind of stuck with what ot gives you. A real person can be instructed much more precisely, and doesn’t have the pitfall of drawing hands awfully or making logical mistakes (multiple/impossible limbs and mangled hands are a frequent issue).

Here, give me a prompt and I’ll demonstrate. I have many credits to spare this month, I’m happy to show you.

By double-posting this prompt.

I was trying out a few different styles/artists. The ones you indicated were in the retrofuturistic style of (the late) Syd Mead.

(The one concession I make is that, out of respect to the living, I only call deceased artists if I’m looking for a particular style.)

Thank you. It…is not an easy road to walk.

If you’re curious, some of the results.

As I mentioned in relation to the parent to this comment, when I received my diagnosis of bipolar disorder, I tried a few prompts about it in Midjourney. They helped me process it (and my dad too, I showed him the results).

Generative art allows more people to communicate with others in ways they couldn’t before, and to inspire and be inspired by others.

When I received my diagnosis of bipolar disorder, I tried a few prompts about it in Midjourney. They helped me process it (and my dad too, I showed him the results).

EDIT: If you’re curious…

Yes. You have to direct the AI. Like, I told it to draw me Stalin as Captain Ahab on the deck of the Pequod (it makes sense in context) and I got this. It was the only of the four options that looked good. But I had to ask it for this specifically.

A human has to supply the prompt. That satisfies the input-of-the-artist criterion. The AI doesn’t randomly gurgle out images; you have to actually ask it what you want to see. (And very often it doesn’t give you what you’re looking for, so you have to get creative.)

There is no functional difference other than the fact that a computer is doing it. It analyzes the common features of the source(s), determines how to square that with the prompt, and generates the image.

We already have machines that invent things for us. A machine that makes art is no different on first principles.

What do you think a prompt is? You’re telling the computer “I want to see this, spit it out.” I can ask Midjourney to draw me the Empire State Building in the style of Van Gogh. It’ll spin the gears and spit out the image requested. (Actually, I’ll do that right now and edit it into the comment. Here it is.) How is that not a derivative work?

RE: Piracy, I’ll quote Sean McGuire, an artist for Marvel:

I keep seeing the argument of “pirate from the giant names” with the implication that it hurts no one, because the company is so big.

I write for Marvel comics.

I write for Marvel comics, my reviews are good, people love the characters I’m writing for, and our illegal downloads are through the roof, because people think piracy hurts no one. It hurts the creators who have fought our whole careers to get through a door that traditionally has only been open for straight white men. Am I the best thing in comics today? Nah. My ego isn’t that big. But I’m good and I’m getting better; give me ten years and I might well be the best thing in comics.

And that doesn’t happen if people pirate my comics and Marvel stops seeing it as worthwhile to give me money.

Transformativity is part of fair use. What AI datasets do is much akin to researchers dissecting corpuses (I’m a linguist, data scraping is what we’ve done in the field for 300 years). When you put something out there, you allow for the possibility that people will see your work and incorporate it into their mental catalog of art and artistic process. Just because a computer does it doesn’t make it magically not OK.

Either information wants to be free or it doesn’t. If you pirate shit then you have no ground to stand on for complaining about AI training.

Meshuggah – “Clockworks”

No worries, I’ll try to succinctly explain (it’s 5:08 AM here and I need sleep).

It’s not so much about efficiency. Calling me “a man with mental illness” trivializes the experience and makes it sound like you can separate the malady from my person. You can’t. Being, as in my case, bipolar and obsessive-compulsive is part and parcel of my life.

“I could care less” is an idiom, a set of words that is somehow beyond the sum of its constituents. It’s apples and oranges with the preceding.

• Person-first language. IT SUCKS. Don’t use it. I am not “a man with mental illness”, I am “a mentally ill man”.

• “X language has no word for Y, so they can’t think about it!” Wrong. Language has ways of adapting to gaps. Sometimes they make up phrases or circumlocute; other times, they straight-up borrow the word. You could perfectly well take some language with 300 speakers in the rainforest in Brazil and talk about quantum physics, though they’d need to invent or borrow quite a bit of vocabulary to do it (but the point is, it can be done in the first place). The Sapir-Whorf hypothesis is basically false in its strong form. (There is evidence for a weak form holding, with particular reference to color terminology and Guguu Yimidhirr’s famed real-space orientation).

• Despite what Twitter and Tumblr tell you, nobody who’s in the demographic actually likes “Latinx” or similar “-x” forms in practice. The people who shout about this ignore that Indo-European languages have had strategies for mixed or indeterminate gender for thousands of years—they would simply use the masculine as a default. The masculine/feminine dichotomy isn’t even thought to have been the original situation in Indo-European: The Anatolian languages evince an older system that was more strictly animate/inanimate, with feminines deriving from some sort of collectivizing suffix (I’m somewhat hazy on the details of this).

• “Can I do X?” “I don’t know, can you?” Piss off! “Can” has acquired a secondary definition relating to permission; this is, cross-linguistically, unsurprising and it’s not clever asking them to rephrase. It’s obnoxious.

• English isn’t the only Anglic language! Off the top of my head I can think of Scots (not Scottish Gaelic, something entirely different) and Yola. Fingallian also appears to be Anglic.

• Arabic and Chinese are often considered single languages, but the situation is akin to Latin and Romance. Modern Standard Arabic and Mandarin are artificial standards, the dialects being highly divergent, so the standards are used to facilitate things generally. (When I was in college I took a few semesters of Arabic and ustaaz, my professor, was from Cairo. He told us once that he was in a room with a Jordanian and a Moroccan and the Moroccan spoke his dialectal Arabic. The Jordanian had to ask ustaaz to translate.)

Yes, it is the Christian ideal of donating 10% of your earnings. In the history of English, “m” or “n” sounds were deleted before fricative consonants (like “s”, “z”, or “th” for example), which is why we say us rather than uns. It’s also why we would normally have said tithe as opposed to tenth. However, on the basis of other numbers (fourth, fifth, sixth), the word for “tenth” was re-formed. Tithe survives in specialized form because that was how it was often used.

Trying to go back to an Ur-language is a fool’s errand. I personally consider it a moot question, though I admit my religious views impact this; even with that said, it’s not really a serious topic of scholarship for many reasons. It’s not even clear that the main major language families share a common origin—a few proposals have attracted serious scholarship (Indo-Uralic, Dené-Yeniseian), but there’s not much in terms of respected research or consensus.

1. Join us on @drums!