Pros and Cons of Generative AI - Debate with an Adobe GenAI Expert
HTML-код
- Опубликовано: 10 июл 2024
- Are you concerned about generative AI tools like Midjourney, Stable Diffusion, Runway and Adobe Firefly taking your job? Is AI a threat to creators or is it a savior?
Let's continue this conversation together in our Community. We have a dedicated "AI for Designers" space, open and free for you to join. All you need to do is to register a free student account: community.yesimadesigner.com/...
=======
IF YOU'D LIKE TO LEARN FROM US:
If you are just starting out, our FREE Graphic Design Survival Kit mini-course will get you going. 8 in-depth topics, 8 video courses. Click here to enroll for free: yesimadesigner.com/survival-k...
ON OUR BLOG
yesimadesigner.com/blog/
FOLLOW US
/ yes_im_a_designer
PODCAST
yesimadesigner.com/podcast/
Apple - podcasts.apple.com/gb/podcast...
Spotify - open.spotify.com/show/2vdeYhJ...
ADOBE CREATIVE CLOUD
If you don’t have a Creative Cloud subscription yet, you can subscribe to it here.
clk.tradedoubler.com/click?p(...
(*We are partners to Adobe. If you purchase through this link you also support our channel.)
==========
DETAILS OF THIS VIDEO
Graphic designers, illustrators, photographers and creators in general are facing a paradigm shift in the creative industry. Generative AI is the greatest disruptor since the invention of the Internet.
Join me and my friend Tareq Fadel, who is an Adobe Generative AI Subject Matter Expert, as we explore this controversial topic.
Work by AI Artists:
app.milanote.com/1QvUrQ1vTR3A...
Midjourney to Runway videos by Useless Editor:
• 100% AI Generated Video
• Midjourney / Runway Ge...
• Midjourney / Runway Ge...
Midjourney photographer and artist reference sheets:
docs.google.com/spreadsheets/...
docs.google.com/spreadsheets/...
Runway ML: runwayml.com/
Midjourney: www.midjourney.com/
Adobe Firefly: firefly.adobe.com/
PhotoShed: www.photoshed.com/
I Am Here | AI Short Film for All-In Summit 2023:
• I Am Here | AI Short F...
ITU AI For Good Global Summit 2023 Press conference:
• ITU AI For Good Global...
Howard Pinsky’s channel: / @howardpinsky
00:00 Introduction
00:31 The impact of Generative AI
04:31 Endless creation of random images
06:57 GAN model
07:46 Governance / Copyright
10:12 Adobe Firefly / Public Domain Images
11:08 Banned Prompt
12:41 Artists / Copyright
15:17 Creative Membership Program
15:59 Impact on Upcoming Creators
20:07 AI apprentices? Future
25:06 Visual Communications / Image Generators
28:14 Outro
Thank you Martin for inviting me and being part of your show. I loved the conversation we had.
That's literally my thesis topic, this video is gonna be very helpful 🫶🏽
Good! It's an interesting topic!
This is very insightful and I think both points can be true during the development of AI
You can’t make full and coherent use of gen AI, without having a good and sound knowledge of Photoshop. You still need to understand layers, selections, masks, smart objects, etc. if you buy a new house and ask a decorator to dress it up according to your wishes, you will still change things to your own liking and resort to other techniques to make your house fit in your idea. Gen AI without knowledge or creativity is lame
I disagree in part to Tariq's opinion about the discouraging effect that AI art will have on young people. Art education in general has already been reduced or cut from public education and there are fewer young adults with basic understanding of art and visual communication. There is a small segment of the population that is driven to create visually that will seek art regardless of the obstacles. There is larger segment that has the potential to create, but can easily be discouraged. AI art is a confusing element for them. Millennials do have a desire to share, but it is an easy, convenient type of sharing that involves taking pics of their meals to share with friends. Yes, it is personal visual expression that was enabled by the convenience of cellphone cameras and social media, but is not comparable to creating art. It is the difference between taking a picture of my clothing for the evening vs designing it. Not the same.
Obviously the Adobe guy would say all that.
Sir I have no PC or laptop.
So I can use Mobile
is that free?
While getting permission and/or compensating artists might sound reasonable when talking about a large company that produces models, it seems very likely to break down when it comes to the ever-expanding range of community-trained models, especially if the optics reverse and you have a large company making demands of a small group or individual. I also would expect a large number of consumers to simply not abide by such restrictions, similar to what happened with music in the early 2000’s.
1:20 I’m noticing that Martin’s voice and face seem a bit “AI”😏
Actually… maybe his video quality was low and he used video as a mocap and used a “meta human” clone from unreal😮
Hi, everyone, I am the original author of the AI video quoted in the discussion: "Useless Editor". I am very happy that the video I made using AI tools has been recognized and liked by Martin Perhiniak. When I have time in the future, I will continue to use AI tools to create some new works, thank you all.
I love your work, I am looking forward to seeing more of your videos! Do you mind sharing the prompt you used for the image I ended up using for the thumbnail of this video? I love those futuristic plastic veiled women shots.
Oh, okay, do you like those pictures? I can email it to you.@@YesImaDesigner
It's the same arguments as the other one... You're using it yourself, but you're also trying to advocate against it, hehe. I'm wondering what you're really doing here tbh... Are you pro-AI but afraid of being cancelled for it so you're gunning for sympathy? Idk... There's a strong dissonance between your actions and what you're saying...
Interesting, legal problem with style is that it can't be copyrighted no more than a musician can copyright a note. What you are referring to would most likely be covered under a design patent, but my style of drawing vs your style of drawing is not unique enough for clear and complete distinction to qualify - again, goes back to the music building blocks. Take "the masters" - where does one stop and the other start? Many people's concept art styles are indistinguishable as they do a functional job. The lines are too blurred and major differences come in with the arrangement of elements, not so much blanket appearance - can't litigate it.
Anyhow, not trying to pick a fight. Came to your other video from LinkedIn. I do however, hope you find whatever side it is that you're actually looking for on this topic. ;)
Thanks for your input. I try to educate myself from every point of view and I find comments like yours valuable. My previous video was made months ago, but I was already cautious and worried about many aspects of gen AI back then, however I tried to stay optimistic. After that video I learnt a lot more about how current models were trained and I felt I needed to make another video about, focusing more on the negative outcomes. I'm still not anti-AI and I still believe there is a place for AI in the creative workflow, but I would never use it commercially in its current form. Gen AI is the most exciting new thing for creators around the world in decades, it's a shame it is overshadowed by the shady practices in the background.
If anything I am trying to position myself as a diplomat working to bring the two camps (anti and pro AI) to agreement and to help achieve a more ethical future for gen AI and creation.
@@YesImaDesigner For sure! That part stands out - you definitely have a solid idea of how it works and what it can be used for. I agree, but ironically that's not my concern with it. My biggest concern is like with the Arabian (?) girl picture situation. When is it gonna hallucinate too close to something that already exists. As far as I know, the only safeguards in place for such thing are fringe definitions and preventative term checking. As a person, you have the discernment to feel: "This feels a bit too familiar" and pull back. AI doesn't have that skill yet.
I'm definitely with you on it's commercial applications - the risks for infringement in the classic sense is a bit high for my liking, hehe. I do use it commercially, but very sparingly, hehe.
@@YesImaDesigner The ethics are only an issue for me where the thing began - the pretense on the use of the training data. If something is for research only, you cannot in good faith produce a commercial product from it, when you know that without that data the product wouldn't exist - that's substantial and highly unethical. I'm glad there are different models - Canva's is great and I love SDxL. Somewhere they found a loophole. There was a moment in Spirit Riding Free where they noted that just because something isn't right, doesn't make it illegal and I think the original training data is such an example.
Well, I'm all for your goal! I'd love for AI to be widely adopted and create new jobs! :D
Thank you for taking the time to respond so thoroughly! :D
There's another artist, Andrew Tischler, who made a video on this topic as well. There were a lot of good discussions in the comments. I'm glad most all artists are all on the same page and are proponents of AI generated art.
I found the video, thank you so much for the suggestion:
ruclips.net/video/M0uADVjOPWg/видео.html
I have been following Andrew, he is an amazing painter. His video about AI is great, he explains the concerns of artists really well.
Uncredited artists have already been producing work under someone else's name