Solve a problem and get rich. This is a problem! A very niche problem and it is even a job! I have done in the past. A tedious mind numbing, boing job but it had to be done! Make images in different sizes for online ads back in the day for flash banners and now for social media ads.
We are here at unter against Ai when it come to Artists work being stolen copyrighted work copied. The blatant theft the stealing the skills, the style, even the brush strokes, the revenue by a handful of prompts typed bing or into mid-journey. There is now a massive class action suite again them for creating the data bank used as the training data set! adcreative.ai but this is great for solving a problem niche where Ai could be a great help. A useful tool.
There are many issues, I tried app.adcreative.ai a few times out of the few times, I tried it i was less then impressed.
There are three sections all are in some way defective but are also unfixable by the client a solution is offered in the end! / first of all this is a MVP a minimal variable product but it has more problems than a dog has fleas! Its really a paid-for BETA! A lot of kinks have to be ironed out!
There are three sub-sections, Text, Image,and Logo, Colour, Font
Finally, the section covering the most problems but doing the least is the section called Logo, Colour, Font, each one has its own unique different problems!
TEXT section
the copy you are to place into the image is too short or could be a bit more elastic, 40 charterers is enough to get my message across, 160 characters would be better! But it would be recommended to keep short as possible - keep it simple!
The Ai text alternatives was not very impressive, the alternative suggestions it came up with well were a bit lame but they all were 40 characters long which i think was more important to you than it is to me. I think this might be due to the AI being forced to keep everything to just 40 characters. This is a case of the tail wagging the dog!
The font size for the "Main headline" renders out not as H1 or H2 but in a smaller size than the punchline, which renders out at! The largest in all my tests. It was the "punchline" that looked like the headline not the main text of all here with me! Then the "call to action" in all my tests is ineffectual in regards and often in the wrong place to where it is placed. A lot of things were in the wrong place. But it's an Ai I expect things in the wrong place space, and I expect a few mistakes but how do I fix them?
I can not I have an uneditable .jpeg
IMAGE Sub-Section
Has its own unique set of problems, The images I uploaded were all 72 DPI images over wi-fi, it took a while for them to upload as there were only 5.
I tried cropping them but it was not possible, to do it quickly, I did not wait as it already took too long, It was close to a minute! All the same, I would have liked at least liked a focus point or a grid the chose the different format sizes, I could choose with each image. I was delighted that I could upload them to the individual library for each Ad.
I tried selecting then all these images to make different background images in my ads with the same text, That didn't work, there was no gif's or transitions offered into the final product OUTPUT. I only to have the first image come out in the all out put images not a variety! Like I expected!
Big problems in little bits - Logo, Color, Font
LOGO
Problems with Logo sub-section It was treated like a cold fish. I uploaded a logo with a transparent background a .png It was just ignored and suddenly had a white background! Not very helpful as I would now have to edit it out.
FONTS
More fonts! there is not man font to selection from.
Suggestions for further improvements!
I would add a fourth segment called style! You can chose the style of ad. you want The 70's 80's 90's or just add a example image, upload or copy&paste a URL, for instance, make my ads it look like an apple ad or post-apocalyptic, like a political poster or the like the current fashion, etc I noticed there was not enough diversity in the results!
What I would do to fix your problem, is have the ads that the customer chooses the customer wants to pay for I would make them editable! Before payment!
OUTPUT results !!
The final product was disappointing, I was hoping for would maybe get 100 different images, and I could use maybe use 5 in ads! The output was, I got the feeling was a bit little lacklustre, I was just a series of stencils that were somehow just filled with the choices that were made in the earlier "Image,and Logo, Color, Font" segment. BUT I must say when I changed the format size to a different image for the final image format sizes! I got all the same mistakes and lacklustre stencil images.
If it had only worked I could save a massive amount of time! Weeks would be cut off my ad creation production time and €€€€ the euros saved would have been substantial! It was just all a little dull, but I found one ad that was ok! just one that was OK. That one ad out in a 100, I would have to paid one credit for because. If it saved me time. I'll pay for it - but I might have to fix it, before I pay for it! I suggest called the DRAFTS segment! Before payment with the credits!
I can't edit the image. So then download it! And I have to recreate it in photoshop! Defeating the purpose or why I came app.adcreative.ai in the first place! I have an E-commerce site with more than 100 plus products. I need some thing that saves me massive amounts of time
100 products x 1 hour to make a stencil = 100 hours times 6 hours to make all stencils in the various sizes, 6 hours from in an 8 hour day 1 email communication, lunch 1 hour handling emails. I have to make ads for almost all of them in a week and place the on various social media! Image sizes! planning and timing! it's weeks of just making ads for each individual product. Add to the time for changes in copy or mistakes!
--
There are many online editors out there CANVA is the big one at the moment, I see it more as a housewife tool and for someone is selling digital products on Etsy. Making Etsy richer but not themselves and gimp the granddaddy of them all it can be programmed to do actions on the server. There is a UNIX server version out there! You could use and photopea https://www.photopea.com/ the best one I have found!
OUTPUT results cond.
So my suggestion to improve app.adcreative.ai because as it now is it is a paid beta. You add an extra stage before the OUTPUT, I'll tell you why later, the output goes into a photopea-like page! Where I can use the "V" key to move text about to suit my needs! and even edit the original copy if I have to then downloads are as .PSD with layers! . Png and or . Jpeg! You do this so the Ai learns and in the future makes better copy and placements of text!
The output was I got the feeling was bit a little lacklustre, I was just a series of stencils that were somehow just filled with the choices that were made in the earlier "Image,and Logo, Color, Font" Section!
A full report to make impovements to adcreative.ai
