Search results

Have you tried Scribus? It is an open source alternative to pagemaker/in Design and it runsin Mac. As for fonts, I'm fond of Century Gothic. It has good legibility and even at 10 it is good on the eyes. For composition, you can't go wring with the classic two-thirds box off-border by two ninths.

You could leave natural 20s without confirmation and confirm others. Or just let it be, are that many crits a bad thing?

Not exactly running out of time. What you're missing is that even though they aren't paying a dime for the material for training, all of these companies are hemorraging money. None of them are actually charging over what it costs to run the huge data farms needed. Open AI just managed to collect...

Prompt engineering will never be a thing. Experience with a model doesn't transfer betwern models (mightn't between versions) and optimizing prompts is something the models themselves can do. No matter how much "skill" the user brings, the results are largely random and unpredictable with most...

Show me one and I'll spread the word.

Please don't, that'd be very degrading.

Only so much. There is a step between the inked -or even just penciled- art and the actual shading/rendering. It is placing the raw color on the canvas so that you can select and mask more easily, and of course have the base color. This part is time consuming and prone to errors -many times...

Sure, most artists would be happy to delegate the fun parts of the process to a machine. In fact modifying a finished rendering with a lot of randomized details -and ghost jpeg artifacting! - is way easier than iterating on a b&w image you had full control of. More seriously, I can only think...

You know that creative commons is a thing right? Also, plenty of people -including yours truly- actually gift custom art from time to time. Besides, these users might even try to do it themselves, perhaps even discovering a related side hobby in the process?

One art youutber once mentioned, a positive to genai is that it filters out cheapskates and stifflers.

Neural networks lack subjectivity and experiences. They can grow and be trained, but what they get cannot be considered "learning". It isn't even remotely the same as a human artist learning from previous art. You probably don't have the experience as an artist or art apprentice to compare, but...

For many, but for clerics and sorcerers no longer having a role-changing/enabling subclass from first level can hurt party composition and dynamics.

Is there a new one? Where can I watch it?

Officially on backlog. I have a lot of stuff to watch pending. Too bad I can't open Prime video without getting distracted by Stargate or Crunchy... And youtube, don't forget youtube.

Back at the turn of the century, I liked Overprotected and Lucky. I didn't know they were autobiographical.

Sorry for being pedantic, but Krashen's is a theory, not a method. A theory explains and informs teaching, an approach guides teaching, and a method prescribes teaching down to materials activities and regimes.

I'm doubtful abou Deepseek, I've heard rumors there's a lot of smoke and mirrors involved. In short, it wasn't as cheap to train and neither is as advanced as originally advertised.

That however doesn't require LLMs. The neural networks that predict protein folding weren't trained by stealing from books.

Training is not piracy, but it is creating a derivative. Which without an explicit license is copyright infringement.

What loss if I bought the album anyway?