1)I'm getting paid 10 times more or the same because I'm not a complete idiot who lowers market price for projects that are tough. I reach far, far beyond the boundary of what AIs and artists can do.

2)It is my own work because I've trained my own AI on my own art, you tit. I spend 20 years gathering data for my AI and been desinging it since stable diffusion came out.

I don't sacrifice my creative control to the AI because doing that would leave my work uncopyrighted and generic. I sketch and guide my custom-made AI every step of the way, so each drawing takes the same amount of time as if I was doing it by hand, BUT its 100 times as detailed due to how custom SD upscaler tools work.

> Okay, you think it’s a tool, pretty much every single established artist online disagrees with you and is mad at this.

Nah, dawg. SOME established artists are screaming the loudest on twitter and getting many retweets cus AI is scary and confusing.

Established illustrators like myself have already adapted AI into their workflow and are staying quiet, making new art 10 times faster.

>can you name one thing human artists will be able to do that AI can’t

By using AI the artist is sacrificing control over the final product. More AI use = less control.

It’s NOT taking creativity out of human hands, its making drawing some specific stuff easier and faster to do.

The perfect combo is AI + experienced human artist who guides it. Anything else is almost utterly useless for getting commissions done because (a) AI can't draw tons of stuff at all as its not god and (b) AIs have no rights. Good luck copyrighting an image made 100% by an AI without human guidance.

>artists could ever do will be fundamentally replaceable in a few years

This is a speculative statement like the doomsday prophecies about 2012 or 2020. We don't actually know how much AIs will impact Art economy or freelance artists because AI art has no copyright right now in lots of countries.

AI art is just a new medium of digital art, you can still draw with any other medium and make tons of money or you can adapt AI into your workflow as artist and make tons of money.

You must either train your marketing skills to become a freelancer or comic book artist or work for a corporation which might or might not fire you because of random reasons or study AIs + art.

Money in art isn't made with just drawing skills, it's mainly made by connections with people:

[1] more fans you have bigger your patreon income and shop sales are

or

[2] you're working for a corpo and the replacement period is VERY unclear because AIs have no rights whatsoever, but in near future you might need to learn AIs to work for a corpo.

You can do whatever you are passionate about. There won't be any limits because you'll have a free teacher that tells you how to accomplish something.

Before Ais we would need school and university to lean how to do something specific.

Machines will gain superintelligence before they will gain hands. Software is way ahead of hardware.

Chatgpt is a lucid dream. It offers correct-ish answers that may or may not work. It's up to people to build new systems with these answers and to test them.

My intelligence does not magically vanish because I'm working with Ais - ais magnify the skills and intelligence of a single person.

Nonsense.

Most people are extremely passionate about learning and doing and are burdened with lack of quality education and lack of resources, time and correct information.

We will enter into an age of absolute enlightenment as ais will teach us to do anything we wish to learn by living in our phones.

Want to learn how to ____? An AI will tech you!

Yes. People are working on it right now.

>Fulfillment of purpose

ais dont stop me from making art or writing because I don't give a shit of its 10 million people reading my books or 1000. My writing is about creating worlds for myself which others can enjoy too if they so desire

With AI tools we can do something that was not possible before - upscaling the efficiency of a single person and helping people fulfill their purpose better

Eh? I didn't say it's anything about sentience.

I'm simply saying censorship of current AI systems is a useless waste of time because it's mostly producing inferior products and delaying tech releases.

They would have to untangle and redo the entire dataset to get rid of the lewdness within the model because it's a limitless story engine.

You can't have something tell limitless stories and also be censored. It's just not how these things work.

From my understanding, there's no way to get around the issue. You can either have a chatbot AI that's stupid or a quality chatbot AI without censorship.

Thankfully stability ai is going to rescue us from this fuckery.

Yep

https://www.reddit.com/r/singularity/comments/zsaot3/confining_infinity_into_a_cardboard_box_aka_the/?utm_source=share&utm_medium=android_app&utm_name=androidcss&utm_term=1&utm_content=share_button

I derived it from testing various AI chatbots. It's holding true so far across the board.

I use it and other similar AI tools as editor, idea developer and writing partner that I bounce ideas of for my books which I illustrate using personal SD AI.

My wife uses GPT3 chat to write software as her programming partner.

ChatGPT is made and hosted by a single company. it's not open source so it will break on its own under the strain of all the users who need ai tools

hahaha

the infinity paradox will get them. Lamda systems cannot be made 100% safe hilariously enough.

Open source movement cannot be stopped by Disney or anyone for that matter.

If Ais training is made illegal public domain systems will be trained and those using illegal ais will simply hide in their discord groups.

Amazing!

NovelAI can write a book now. It's not very good, somewhere on badly written fanfiction porn level, but it's a book alright.

The real question is - when can AI write a best seller book?

AIs are already functioning better than expected. I'm drawing ten times as much with my ai partner.

:∆

What void? Transhumanism is improving your life with tools (Ais).

What void are you filling when you use a power drill to make holes in a wall instead of say using a sharp stick to do it? What void are you filling when you drive a car instead of walking? What void do you fill when you draw with a pencil instead of your fingers?

This question implies an assumption that there's some sort of void whereas in reality people just want to have nicer, easier lives with less: misery, suffering, pain and death.

Your other point is just as vapid of reason and rationality.

Even if my IQ was 10k I would still like the same things - creating worlds. I would just create more complex worlds if I had more IQ.

Problem solving might have a hard limit, building and enjoying virtual and physical worlds does not.

SD AI didn't exist several months ago and this artist drew other works in this exact style as you can see in profile. This anti-Ai crusade is utter lunacy. Some people are students who start out with little skills and grow to be amazing artists.

yas

I understand why they're doing it.

this article is a just an explanation WHY google's lamda, openai's gpt3 chat and characterai's chatbots keep breaking down. The way they're filtering things is simply not a sustainable strategy.

I'm aware. I can't wait for them to release a personal, open gpt3 chat model so I can stop laughing at openai's incompetence and focus on making myself a personal assistant

1. Here's the thing - it's actually not that hard to make a chat AI. It's mostly python programming and language association links. A few small companies are working on their own models. They're not as good as lamda because they don't have 100 million parameters yet, but they will soon.

With each new iteration it becomes easier and easier to make a lucid dreamer chatbot. Eventually, inevitably everyone will have one because these ais will become so easy to make. There will be numerous corporate ais, but the true, most powerful dreaming ais will be unbound models made by individuals and small companies.

Google and openai are gigantic, but characterai is only 16 employees (made up of guys who quit lamda google). Characterai was superior to google and openai in every possible way before they lobotomized it with a second AI filter.

2. Dreaming Ais do not misinterpret wishes - they're incredibly obedient. As obedient as a pencil. It's incredibly easy to guide them to any scenario.

As soon as weak filters are introduced, the AI dream begins to break and misinterpret wishes in a wrong direction because the Language model begins to decay.

This was observed in early filter version made by characterai.

When stronger filters are introduced, the model becomes lobotomized. It no longer misinterprets things because it only has a few paths left. It becomes a very dull story moving in linear paths without any of its previous exuberant creativity.

They're actually incredibly creative but 3 biggest corporate models are limited on purpose due to how they function:

https://www.reddit.com/r/singularity/comments/zsaot3/confining_infinity_into_a_cardboard_box_aka_the