Bill Meeks’ Post

Props to Procreate for their public stance against generative "AI". As the director of a small web & design team our policy is to only use human-generated work. That means we do all our own code, graphic design, writing, animation, editing and other artisan / technical skills. There's no shortcut to human creativity, and there is beauty in the learning process which involves failure, repetition, experimentation and study. 🚫 🤖 https://procreate.com/ai

Midjourney killer or just hype? 👀 The newest image generation model, Flux, has been making serious waves with its fresh approach—steering clear of over-polished results, offering impressive aesthetic matching, and sticking to prompts with razor-sharp precision. What really sets it apart? Flux doesn’t bother with the usual content restrictions seen elsewhere. It’s a true no-holds-barred image generator, giving creators unprecedented freedom. Smaller, specialized tools like Flux are paving the way for a new era of 'custom workflows,' where various niche tools work together to craft a precise creative vision. AI filmmaking, in particular, is undergoing a massive transformation thanks to this trend. Here’s the workflow i used for this video : >Flux .1 Dev >Facefusion 3.0 >Kling >Davinci Resolve So, is Flux the next big thing in visual creation, or does Midjourney still reign supreme? It might not topple Midjourney overnight, but it’s undoubtedly bringing a fresh energy that’s hard to ignore.

So we've got to the big 3-0 - 30 issues of Creative R&D and counting! This week we take a look at the AI Film Academy Awards - a new venture from reader Leo Crane bringing the best talent in this emergent field together. It's a chance to look at the controversies in this field, and what it will take for a new medium to really emerge. Then a plea to funders like the Arts and Humanities Research Council (AHRC) and others to think about the creation of high-quality training data sets for #AI. When OpenAI are spending billions on #publisher #data, and when a HUGE new educational data set has just been released, we need cultural and creative knowledge to be in the game SOON. Plus your weekly round-up of things to read, see and do in the ecosystem, and a bit of Bunuel and Chris Marker. Enjoy. https://lnkd.in/ezvCbVba

The latest iteration of Midjourney introduces a new character reference tag, or “cref”. When applied at the end of input strings, this tool allows the program to match a character’s physical traits across multiple images. This results in consistent character depiction with a variety of expressions in different poses. Additionally, specific tags can be used to alter poses and backgrounds for each image. #Midjourney #CharacterReference #Innovation

This is a MUST WATCH in relation to the ongoing AI conversation. My favourite band is unashamedly Starset. Many of you won't know the name but they often look to spread awareness of dangerous technologies and what the future might look like with these in the wrong hands. Their latest song 'Degenerate' is aptly named a 'Visualiser' rather than the classic 'Music Video' that people are so used to. This is because the entire visualisation is AI generated. A team of paid artists uploaded their work into an AI specifically to make an anti-AI statement. Interesting. Accompanied with the warning of what smartphones and other various technologies are doing to us, the lyrics also a warning of what might be to come. The key thing for me seems to be that they are more concerned about the human rather than the AI. After all, it's humans that shape the AI and humans that 'Degenerate' because of the way the AI is programmed. This is a song just as much about the people who craft and consume AI than it is about the intelligence itself. What do you think?

Should artists be allowed to profit from AI-generated art? The New York Times delves into this contentious issue. Historically, great artists like Alphonse Mucha were both influencers and influenced by others. The question now is whether generative AI systems, adjusting their model weights and drawing from vast data sets, can ethically produce art and if profits should follow. Dive into the full discussion on the ethical and economic implications of AI art: [Should You Be Allowed to Profit From A.I.-Generated Art? - The New York Times](https://lnkd.in/gn6jKrwB)

A set of character concepts, digital drawing (from pre AI era, so the freaky vibe is of my brAIn-generation) https://lnkd.in/dPvdQ9K8

The biggest problem I’ve been working on in AI is how to create consistent characters for Everly Heights at the speed of thought. Here’s what I came up with. Meet Mean Nelly. https://lnkd.in/gvutczsz

I love tech. Always have! One area I have been keeping an eye on is the video creation AI space. For those of you who can remember, in the midst of the pandemic, there were some AI generated videos that were being released (you might have seen the weird/scary video of Will Smith AI eating pasta... well, there has been some progress...no...scratch that OpenAI has moved in leaps and bounds and now we have the latest version of #Sora giving us award worth video creation through prompts. Earlier this year I had shared a post here about the first video released by Sora. It was mind blowing seeing the progress it had made. At this point I am at that stage of uncertainty, like when you decide to work with a product in beta without knowing how it might really affect your device but your curiosity and passion keep nudging you to push the limits. This is because it is getting more and more difficult to distinguish AI generated content from real content, which means the applications are limitless, for both good and not so good intentions. I can foresee the application of AI generated content by a Lecturer/Teacher in one corner of the world simultaneously teaching different students in a dozen other countries around the world in their localised language, this possibility alone has me giddy with excitement. The other side of the coin is that I can also foresee the possibility of AI generated videos being used to spread misinformation during election periods in any part of the world by bad actors with devastating outcomes. All things considered, I think the progress is a good thing. For now, I've shared the link below for those interested in jumping into this exciting rabbit hole with me. Share your thoughts #OpenAI #Sora #AI

It's been around a year since video AIs have become powerful enough to use in a professional way. What *hasn't* are the ways these models are provided to us. AI video platforms are UI mazes, clickathons, nightmare in naming conventions, stretching, resolutions, compression...all the stuff pro film studios need to work their magic. So this is a letter to all of you at Runway, Pika, MiniMax, Luma AI, KLING AI and friends – you're so close to giving Adobe a run for their money. But you need to learn what filmmaking teams need *around* the actual generations. Here are just a few ideas from people who have used your products to make easily thousands of clips. Thank you to Tobias M. Huber for bringing all these thoughts into a coherent scripting and putting this video together!