Dallas, I looove how honest and useful this is. I watched you go through this in real time and was rooting for you the whole way. All three systems Claude, Flux, and the watercolour itself need creative room simultaneously. You had to break it to see it.
And the fact that the answer arrived at the kitchen sink not during one of the many hours you spent in Claude fighting prompts is very representative of how breakthroughs actually work.
The navigator is back where she belongs. Worth every frustrated credit. 🩷
It's been a wild ride that's for sure! 🤯I could not believe my brand new brand was breaking! The thought of trying to find another visual identity just sat terribly wrong so I had to fight for this... and blow an awful lot of credits and time, lol.
Funny how the answer usually arrives when we just stop and do something else, and incredibly relieved to now move on to a new focus 😂
You were there since the very start! Incredibly grateful for your guidance and kind support at every step! 🩷🩷
I know the feeling. It’s like an existential vacuum. I had a similar scare a few months ago when my tool started charging way too much for the same features I used for months. I always appreciate how reflective you are and your brain’s ability to observe the meta layer. I am learning so much from you🩷🦩
P.S. Claude has convinced me it is entirely possibly to build my own little image tool that will reduce costs significantly - that part is the worst when you are onto something that works! Will see how that goes... watch this space 😂
This was such a raw post :) thank you for this! Especially because I feel it to my bones too. I LOVE to add rules after rules until the stuff that was working stopped working and I end up not knowing if I broke it or not??? Haha.
I think the answer is always both :) We should strip more stuff back to its original version...
It's such a delicate issue I think because sometimes we absolutely need lots of rules, other times, we need to set a clear output requirement and leave the rest with a light touch... figuring out which one is needed when and how/why is HARD! I could not let this particular issue go until I had it fully figured out, took the hard route, but definitely learned things I could not have learned the easy way around.
I'm a perfectionist at my core and I laugh at how I chose a medium that I cannot fully perfect or control - was Past Dallas thinking anything beyond "ooh pretty pictures"?! 😂
I've experienced the frustration of images not generating after they have been... yowza. It's painful! Nothing like breaking things... multiple times, to get the clarity we need.
Way to stick with it, my friend. I don't know that I would have connected the texture of the paper and the nuance of watercolors!
Thanks Kim! I could not believe that I had literally just figured out my visual story and was then struggling to pull it off 🤯😅 I am now in the "can laugh about it" phase, lol, because the arc is hilarious. Nothing like taking the hard path, but I did learn a ton!!
I spent wayyyy too many hours today working on my images. And do you upload the GIF file inside your post? I see it working here. I’ve just been doing the thumbnail, but would love to do the full GIF— but didn’t think it worked? Am I missing something? (See, I can’t stop! Hahaha)
Hey Sharon! Yes, here in this one I uploaded the gif into the post itself. It takes a little bit to load and I admit to a minor freakout when my post published and the gif looked like it was missing - it appeared about 30 seconds later so guessing it just needed time to sort itself out! You can also upload MP4s videos into posts which I usually do if I don't need to them to appear as a thumb nail as well.
Good luck! Adding animation to posts is addictive 😍
I’m enjoying it! But maybe I didn’t wait long enough last time! The MP4’s have to have a play button, right? Like that your GIF just plays in the post without having to hit play. Enjoying it! And thanks for the tip, really helpful.
Yip, correct - that is why I chose a gif this time around too (I forgot that part until you said this!) as it sits right at the top and the play button looked lame!
I love your stories and your learning and your ah-ha moments!
And I’ve been there trying to program an AI system and as I constrain it more and more, its output gets worse. It’s one reason why I completely dropped off of custom GPTs!
And why I prefer to program as much as possible instructions into code and context files for processes that have to go the same way every time.
It’s also why I don’t use long structured prompts. I just talk to the AI in short bursts to calibrate alongside it.
I remember when you were telling me about your artifacts breaking down and learning about system limitations vs the tool! It’s so cool to see the resolution at the end of it.
You've been around witnessing it, Alyssa, and always cheering me on! Thank you!
It's a fascinating journey in amongst the frustrations?! Constraints are funny things, they work until they just don't. My image prompts had crept up to around 3,000 characters from around 1,000 😂 I'm with you on not liking long structured prompts (aside from what image prompts became, lol), I tend to prefer working in layers which helps with that calibration too.
It is a fascinating journey! And often because of the frustrations (even if it doesn’t feel that way in the moment…)
I love how we can laugh about them when they don’t work. That’s what I also appreciate about AI, is how it can be fun when it works and doesn’t work.
And even when it’s frustrating, the number of times that I’m angry at it and then suddenly there’s that insight about why it was being dumb haha it just is so funny to me, and lights me up. And then that’s another thing to learn about how this random technology works!
Getting that insight into why it is being dumb is everything on a bad AI day!! Lol. I love those moments though because the "ahhhhh!" sort of makes it all worth it and you don't hit that issue again.
Oh Dallas ... how I resonated with this article. I can't tell you how many times I've thought things are my inexperience rather than a platform problem. Everything keeps changing and we need to continually adapt, and sometimes its hard to sense what the issue is.
AND THIS IS SO REAL, I loved this line --> "I could not have arrived at the simple answer without exhausting every complicated one first."
I think that says so much about who you are. You're determined, not willing to give up, but also aware enough to know when to sit back and let it breath and let it come to you.
Thank you, Natalie! You said a line to me ages ago about how you'd learned that AI animation seems to work best with mood and motion rather than set constraints, and it just stuck with me. I think it is so true, in my experience anyway.
Learning to tell if it is "me" problem or the tool's limitations, or whether the platform has issues is no easy feat 😂
I think what drives me is that I don't like not understanding a process and I really don't like accepting "good enough" and just moving on. It does get a little obsessive along the way, but I think it's good because it means I get somewhere... eventually, lol.
FANTASTIC, Dallas! I got more from your story about your square dance with your imagination + AI + image generation + human trust.
My eyes bleed from all of the articles with pages and pages of screenshots of someone's AI adventures, all trying to say, "Hey, look at me! I'm riding the bicycle without training wheels." Still not wearing a helmet and still not noticing the traffic in the street, but still riding that bike, right?
Just in the past 6 months, there have been so many writers using some kind of AI thingie with the watercolor thingie. I'm still placing my bet on the writer who knows how to tell a story and struggles to write. That's my favorite "tell" with writers these days. Their words do not come easy and editing is obvious, plus the words and the story connect with in the same way great writing did before ChatGPT.
Thanks Georgia! So glad this resonated with you! It's been quite a journey, I feel like I have reached a conclusion, but there is still part of me that also believes this is possibly only part one - AI has more curve balls to throw my way and things to learn 😄
I so relate to this, Dallas. I built a repurposing skill in Claude where it gives me notes, linkedin posts and 2-3 prompt images based on the article. I let the AI draft the prompt but refuse to touch a single word until I see the first render. It’s a strange dance of trusting the machine’s 'vision' while keeping our hands on the emergency brake. Sometimes I'm happy, sometimes I'm like wth? I feel like once a while we need a refresh because the tools are changing and we need to fine tune our processes.
There is something really comforting in realising it's not a unique problem in this AI adoption era, hey! I'm still learning where to restrain tightly and where to remain in the loop but just point towards the outcome and trust the process.
Definitely agree that processes need refreshing - I now have a strategy for what to do when this happens again (not if, because it feels kinda guaranteed with the way models constantly change!).
Dallas, I looove how honest and useful this is. I watched you go through this in real time and was rooting for you the whole way. All three systems Claude, Flux, and the watercolour itself need creative room simultaneously. You had to break it to see it.
And the fact that the answer arrived at the kitchen sink not during one of the many hours you spent in Claude fighting prompts is very representative of how breakthroughs actually work.
The navigator is back where she belongs. Worth every frustrated credit. 🩷
It's been a wild ride that's for sure! 🤯I could not believe my brand new brand was breaking! The thought of trying to find another visual identity just sat terribly wrong so I had to fight for this... and blow an awful lot of credits and time, lol.
Funny how the answer usually arrives when we just stop and do something else, and incredibly relieved to now move on to a new focus 😂
You were there since the very start! Incredibly grateful for your guidance and kind support at every step! 🩷🩷
I know the feeling. It’s like an existential vacuum. I had a similar scare a few months ago when my tool started charging way too much for the same features I used for months. I always appreciate how reflective you are and your brain’s ability to observe the meta layer. I am learning so much from you🩷🦩
Aww thank you so much! Right back at you! 🩷🩷
P.S. Claude has convinced me it is entirely possibly to build my own little image tool that will reduce costs significantly - that part is the worst when you are onto something that works! Will see how that goes... watch this space 😂
Welcome back love! ❤️
This was such a raw post :) thank you for this! Especially because I feel it to my bones too. I LOVE to add rules after rules until the stuff that was working stopped working and I end up not knowing if I broke it or not??? Haha.
I think the answer is always both :) We should strip more stuff back to its original version...
Thanks Mia!! Back at it 🩷
It's such a delicate issue I think because sometimes we absolutely need lots of rules, other times, we need to set a clear output requirement and leave the rest with a light touch... figuring out which one is needed when and how/why is HARD! I could not let this particular issue go until I had it fully figured out, took the hard route, but definitely learned things I could not have learned the easy way around.
I'm a perfectionist at my core and I laugh at how I chose a medium that I cannot fully perfect or control - was Past Dallas thinking anything beyond "ooh pretty pictures"?! 😂
youuuu and me both! Past Dallas and Past Mia were definitely not thinking about what they were getting themselves into!! 🤣
Future Mia and Future Dallas are hopefully very grateful though 😂
I've experienced the frustration of images not generating after they have been... yowza. It's painful! Nothing like breaking things... multiple times, to get the clarity we need.
Way to stick with it, my friend. I don't know that I would have connected the texture of the paper and the nuance of watercolors!
I love your images!
Thanks Kim! I could not believe that I had literally just figured out my visual story and was then struggling to pull it off 🤯😅 I am now in the "can laugh about it" phase, lol, because the arc is hilarious. Nothing like taking the hard path, but I did learn a ton!!
Haha… girl, I often take the hard path myself! At least we have good stories to tell!
I spent wayyyy too many hours today working on my images. And do you upload the GIF file inside your post? I see it working here. I’ve just been doing the thumbnail, but would love to do the full GIF— but didn’t think it worked? Am I missing something? (See, I can’t stop! Hahaha)
Hey Sharon! Yes, here in this one I uploaded the gif into the post itself. It takes a little bit to load and I admit to a minor freakout when my post published and the gif looked like it was missing - it appeared about 30 seconds later so guessing it just needed time to sort itself out! You can also upload MP4s videos into posts which I usually do if I don't need to them to appear as a thumb nail as well.
Good luck! Adding animation to posts is addictive 😍
I’m enjoying it! But maybe I didn’t wait long enough last time! The MP4’s have to have a play button, right? Like that your GIF just plays in the post without having to hit play. Enjoying it! And thanks for the tip, really helpful.
Yip, correct - that is why I chose a gif this time around too (I forgot that part until you said this!) as it sits right at the top and the play button looked lame!
Well this looks amazing, you nailed it!
Thank you so much! 🩷 I love it too.
I love your stories and your learning and your ah-ha moments!
And I’ve been there trying to program an AI system and as I constrain it more and more, its output gets worse. It’s one reason why I completely dropped off of custom GPTs!
And why I prefer to program as much as possible instructions into code and context files for processes that have to go the same way every time.
It’s also why I don’t use long structured prompts. I just talk to the AI in short bursts to calibrate alongside it.
I remember when you were telling me about your artifacts breaking down and learning about system limitations vs the tool! It’s so cool to see the resolution at the end of it.
You've been around witnessing it, Alyssa, and always cheering me on! Thank you!
It's a fascinating journey in amongst the frustrations?! Constraints are funny things, they work until they just don't. My image prompts had crept up to around 3,000 characters from around 1,000 😂 I'm with you on not liking long structured prompts (aside from what image prompts became, lol), I tend to prefer working in layers which helps with that calibration too.
It is a fascinating journey! And often because of the frustrations (even if it doesn’t feel that way in the moment…)
I love how we can laugh about them when they don’t work. That’s what I also appreciate about AI, is how it can be fun when it works and doesn’t work.
And even when it’s frustrating, the number of times that I’m angry at it and then suddenly there’s that insight about why it was being dumb haha it just is so funny to me, and lights me up. And then that’s another thing to learn about how this random technology works!
Getting that insight into why it is being dumb is everything on a bad AI day!! Lol. I love those moments though because the "ahhhhh!" sort of makes it all worth it and you don't hit that issue again.
Yes!! Exactly! I don’t care if it’s wrong I just want to know whyyyy and then I can do something to fix it.
Oh Dallas ... how I resonated with this article. I can't tell you how many times I've thought things are my inexperience rather than a platform problem. Everything keeps changing and we need to continually adapt, and sometimes its hard to sense what the issue is.
AND THIS IS SO REAL, I loved this line --> "I could not have arrived at the simple answer without exhausting every complicated one first."
I think that says so much about who you are. You're determined, not willing to give up, but also aware enough to know when to sit back and let it breath and let it come to you.
Thank you, Natalie! You said a line to me ages ago about how you'd learned that AI animation seems to work best with mood and motion rather than set constraints, and it just stuck with me. I think it is so true, in my experience anyway.
Learning to tell if it is "me" problem or the tool's limitations, or whether the platform has issues is no easy feat 😂
I think what drives me is that I don't like not understanding a process and I really don't like accepting "good enough" and just moving on. It does get a little obsessive along the way, but I think it's good because it means I get somewhere... eventually, lol.
I think we are very similar! 🩷
I think we are too Dallas! I always want to know how and why things work, not just that they do or don’t.
I think it’s an important curiosity to have! To question these things ❤️
FANTASTIC, Dallas! I got more from your story about your square dance with your imagination + AI + image generation + human trust.
My eyes bleed from all of the articles with pages and pages of screenshots of someone's AI adventures, all trying to say, "Hey, look at me! I'm riding the bicycle without training wheels." Still not wearing a helmet and still not noticing the traffic in the street, but still riding that bike, right?
Just in the past 6 months, there have been so many writers using some kind of AI thingie with the watercolor thingie. I'm still placing my bet on the writer who knows how to tell a story and struggles to write. That's my favorite "tell" with writers these days. Their words do not come easy and editing is obvious, plus the words and the story connect with in the same way great writing did before ChatGPT.
Thanks Georgia! So glad this resonated with you! It's been quite a journey, I feel like I have reached a conclusion, but there is still part of me that also believes this is possibly only part one - AI has more curve balls to throw my way and things to learn 😄
I so relate to this, Dallas. I built a repurposing skill in Claude where it gives me notes, linkedin posts and 2-3 prompt images based on the article. I let the AI draft the prompt but refuse to touch a single word until I see the first render. It’s a strange dance of trusting the machine’s 'vision' while keeping our hands on the emergency brake. Sometimes I'm happy, sometimes I'm like wth? I feel like once a while we need a refresh because the tools are changing and we need to fine tune our processes.
There is something really comforting in realising it's not a unique problem in this AI adoption era, hey! I'm still learning where to restrain tightly and where to remain in the loop but just point towards the outcome and trust the process.
Definitely agree that processes need refreshing - I now have a strategy for what to do when this happens again (not if, because it feels kinda guaranteed with the way models constantly change!).