So cool,
Look what I did with this build
Quite interesting, especially how it changed the colours, added shadows and highlights based on the sun angle. Did it automatically add the tie down chains and such as well?
As cool as this is, there is a part of me that is unnerved and slightly terrified about what this means in terms of future and historical information. The fact that it was so easy and accessible to change the entire context and perspective of the original picture gives one something to seriously think about.
All in all, a very well done Super Hornet. Would be nice to be able to print out the background and deck to display.
I didn’t do anything but upload my pic and select a setting.Yes it could be scary inthe wrong hands.
This is how I figure out the colors of my auto builds, I just tell AI what kit I have and tell it to show me in red or black or whatever, it creates the picture, then I can add details like change the hood color or whatever. It’s pretty neat. It will also color match if you need. “ I need a Tamiya acrylic match for Revell color 123” or “ using Tamiya acrylics how do I match Chevy engine red” and it will tell you what colors to mix and the ratios.
No,but it is fun,and ego inflating.
Expect to continue to see my normal stuff.
Yours looks awesome too.
Ok the AI version is way better, but my god the background generation is very cool! It even turned my original photo around which is kinda strange and put 1 pilot in front, but goose is missing from the rear seat? I feel like it used my photo to just find a real photo of the black lions F14 and used that for it’s AI generation?
Do I dare ask what the link is to the sorcery?
Happy Modeling,
Mike
It’s right on the top of my initial post
Oh geez, thank you… I thought that that was an advertisement at first ![]()
![]()
Happy Modeling,
Mike
Ok, so I tried sorcery and its wonky. I did a few with “add miniatures” and the pics I added, are not the outcome pics lol
They even gave my Scooter ordinance and added red paint to the slat housing ![]()
The first pic was of a front three quater few and the otner pic, was an upclose pic of nose/office..
This one, they added a pilot which is coo.
I did this one of my P-40.
The last one, I.did of an aerial view.. Neat but its a drone lol
Wonder when they decide when a pilot is needed or not?
Over all neat but I too have mixed feelings in regards the extent and usage of AI.
Happy Modeling,
Mike
Ok, now these are pretty cool but when you look at it closely, they adapted my markings on a real Stang.. Review mirror and seat belts are a few give aways. Still neat
My Tamiya Porsche..
And my Revell 41 Willys… Thanks @Tojo72 ![]()
Happy Modeling,
Mike
It’s amazing how far technology has come along. You’d swear all of them are real. My two favorites F-14, and P-51​:+1:![]()
Like I mentioned, I do not fancy AI for many reasons.
Case in point. My 1/35th armored Jeep.
Now I tried the “add figures” and I believe the majority of the ppl here know that figure painting is my kryptonite..
With that being said..
Yeah,really just for fun,what would be the point of making people think I was better then I really was
This is already becoming a problem in the figure-painting social media space. It’s trivial to take a photo of an unpainted mini and ask AI to “paint” it. Then you can create an Instagram account with nothing but top-tier, impossibly-good, painted figures, rack up thousands of followers, and the minipainting hobby is forever diluted and tarnished, because nobody can tell what’s real and what’s not. Some of the images in this thread could be passed off as world-class dioramas – the personnel in @Butz’s middle P-51B airfield image for example, look like expertly-painted figures, not real people.
I’m also seeing it on history pages, and it cuts in both directions. There are entirely convincing AI-gen black and white WWII-style images of events that didn’t happen. At the same time, real images that I’ve seen for decades are being labeled as "AI’ by people who have had their credulity poisoned, have no knowledge of the time period, or even have a political axe to grind.
For example: how do we plan to accurately find out the number of rivets on the canopy frame of a FW-190 D-9 when you can’t tell if the reference image is a real photo of an FW-190-D9, or an AI generated image? Already the images in this thread are probably amalgamated into some AI-training data set, and are further watering down the “internet’s idea” of what the real places, events, vehicles, people, looked like, what models look like, what painted military miniatures look like.
Books are great for now, but there will come a day when AI-generated images find their way into print reference material. Maybe it’s already happened. We’ll have to start including caveats on our displays: “all cited reference material was hard-copy, printed before 2020.”
I definitely can see the fun and novelty of putting our builds into a “realistic” setting. It’s astonishing how well it works and has a very high cool factor. It brings our imagination to life with a couple clicks. But it sure makes me uncomfortable.
Maybe it’s the journalist/photographer in me being a curmudgeon. But I’ve just started delving into the diorama/storytelling side of modelmaking – it’s an art form that takes decades to master. This all feels a little like a middle finger to Sheperd Paine’s legacy.
I can see a counter-argument to this: that these are just digital images, and the true value of building a model is having the physical object in your hand or on your shelf, enjoying the experience of the build, and sharing in-person at club meetings and shows. I accept that.
At the same time, if there’s no inherent value in sharing accurate digital images of our work, then the purpose of this forum starts to become hazy.
The “dead internet”, where nothing is real, or can be proven to be real, is already upon us in many ways. But we’re fighting it by keeping forums like this one alive! FSM and similar spaces are bastions against the dead internet. While most of social media is rife with bots and AI, this is still a place with real people on both sides of the screen.
Forums like this are the closest thing to an in-person modeling community the Internet has to offer. Let’s keep it real. ![]()
Well said Tom! When it comes to doing any sort of research, I go back to my old trusty Squadron In Action, Walkaround, Verlindens Lock On and other such books.
I laugh when ppl say that that type of research will be out dated but I have a feeling AI will bring us back to the ol school books.
As for the Mustang pics, the first photo, setting is as real as it comes including the personal!
But look at those Mustangs again. That’s a real Mustang with my paint scheme/markings on it.
I’m an ol school modeler so AI wont last in my vocab!
Happy Modeling,
Mike
Why fake? AI only sets your model where it should actually belong without the fuzz to have to build an entire diorama and at the same time also gives you hints on how to build one if your intention were to actually do so.
In my humble view quite useful indeed.
And yes, in both cases I liked the setting very much.
Not exactly sure I said fake but. For my Jeep for instance, I added “minature figures” and what looks like Alpine Figures appeared.
I mentioned that a builder could use said pic with crew and say that they did the Jeep along with the figs. How would we know if they did it or not?
For my Mustang, the photos are not actually of my buiid but of a real aircraft. Look at the seat belts, rearview mirror, ties, landing light ect.
Yet, the photos of my Porsche, weird enough are of my actual build.
Yes, I see your point about using it to have an idea of a possible diorama along with ideas but, how about those using it to stretch the truth about their build?
Hence my example with my Jeep with the figs, to pass it off as their work.
Besides, I thought this particular AI program was supposed to use the photo you picked out and not to pick real and model?
Happy Modeling,
Mike
Toimi_Tom, you are right of course and your considerations are well motivated, but I dare to say that nevertheless, at least for those of us who know their history, its events and actual locations and details (needs a lot of serious research and study), such AI created images offer only a hint to what one might potentially do with either a vignette or a diorama.
It offers an idea, not reality in itself.
It is then up to the individual to go deeper into the subject and choose wisely an appropriate setting to build for his finished model.
The same applues for the models and figures themselves.
Those who jump the caboose due to an AI reproduction or because of it are just people who don’t know better.
This is why I have always advocated for an appropriate education in schools in which teachers actually help their pupils to identify and differentiate fake from fact.
It is not more complicate than to distinguish between an advertised product and its true effectiveness, or separate propaganda from true reporting, or again narrative with true facts.
Unfortunately lots of people have only learned to trust artificial means of communication due to the complete lack and thorough introduction to these artificial means, whether PCs, laptops or other devices.
This is also why so many people suffer depression feeling lonely. They have voluntarily isolated themselves by trusting only mechanical devices and don’t know how to master them appropriately.
It is not just an educational matter but indeed a mastering one.
Those devices, including AI, can easily be controlled and circumvented instead of being dominated by them if we only returned to use our own minds in a more active way and returned to communicate with our fellow human partners and friends on a one on one basis, instead of always transmitting remote verbal or written messages.
Electronics have by now transformed a lot of people into drones, into unwilling and unknowing slaves and it didn’t need to be so if someone had had the good idea to appropriately educate and train these people on how correctly to use them.
But this goes against the grain and interests of those industries that produce such devices, because should people learn their appropriate value, there wouldn’t be such a prosperous influx of money into them.
And yet, many people also allow themselves to be lured into this trap, since they think it is more convenient to be ruled rather than being independent, either because they don’t know how or are unwilling to make an extra effort to develop further their own given intellect.
In my humble view the process could easily be reversed if those who have, or should have, the responsibility of our general well being, did impose certain fixed rules and regulations about the use of these artificial means.
After all, one cannot drive a car or a truck without the appropriate driver’s license, and if caught without it they get either punished with a hefty fine or even deprived of the right to drive their own vehicle.
And how does one get a driver’s license? By studying the laws of conduct on the road and in traffic, as well as mastering his or her vehicle with prudence and care. Without these two factors one does not get such a license.
So why not do the same with iPhones and other electronic devices, including how to handle AI?
It is very possible and I dare say by now, more than imperative.
Fortunately for me, I come from the entertainment business and once in it one learns fast what is real and what is not, of course if appropriately trained to do so. There too one must undergo years of study and research, and after all, what many do not understand, this is also an industry, a very serious and highly competitive one, and only those who actually have worked in it know the tricks, and sometimes also the sleazy ones, involving such a discipline and therefore have learned to avoid unpleasant surprises coming from it.
After all what is AI if not fictional intelligence? And as such it should really be regarded, no matter how smart or indeed, intelligent it might seem.
What we still fail to realize is that AI is just the sum of the knowledge that their creators have allotted to it, and in many cases it is still lacking a true and active thinking and fails miserably if the question posed to it eludes its common knowledge.
Some fear it “coming alive” but to come alive it should also have the human experience incorporated in it, with its high and lows included, a thing apparently that AI does not fully understand nor compute.
To have those AI should first be organic and it is not. It is just a tool, even if in the form of a highly sophisticated android like those produced in China or Japan.
As I always say, if it doesn’t bleed it ain’t human, or as Data would say: “If you pinch me do I not leak?”.
We do not have to be afraid of AI, but we need to be aware and alert, and because we are human, we can and must be. A thing no machine is really capable (as yet) of doing.
Also, and as far as I know, no machine can eat or drink and taste food or drinks. It can understand their chemical components but cannot taste them.
And therefore to keep it finally short, we do not need to be afraid, at least not here in this forum, of AI, since we all have still very functioning and creative minds, otherwise we here wouldn’t produce so much stuff as we do.
Sometimes even more than one at a time just for variety.
And speaking of variety, that’s another quality AI doesn’t have.
I know that a photographer is particularly sensitive on what AI can produce, but as I said before, the critical mind recognizes a fake from the original one almost immediately just because we also have an instinct in such things and if a picture looks too perfect it probably is either a fake or a retouched one.
The same science that has produced AI has also produced devices and apps that can effectively determine the difference between the two, but mainly our experience and our studies and research drive us to smell a rat when there’s one.
There will always be nincompoops that have something to say about the “fake” moon landings and they will always pester us with them until we will take them on a Moon mission and let them touch the landers with their own hands.
Wanna bet? Even then they might not be convinced and will say that they were built there expressly for this purpose, by even imagining aliens having done so.
So what can I say other than let them think “elephant” if they want to.
Those who know better will always know what is and what is not.
Don’t you agree?
Let me know and let’s discuss it.