• ℹ️ Heads up...

    This is a popular topic that is fast moving Guest - before posting, please ensure that you check out the first post in the topic for a quick reminder of guidelines, and importantly a summary of the known facts and information so far. Thanks.

2025: General Discussion

Certainly is great to see another day added, but there's an elephant in the room here that needs to be addressed...

That crowd very much looks like it's AI Generated. If so, it's such a shame to see them stoop to such an egregious low. I shouldn't have expected less from the park at this point, but there we are. As long as they don't use any of these 'tools' on park in any attractions, otherwise we've a MUCH bigger issue on our hands.

1749568028553.png
 
What exactly is the issue with the park using generative AI tools, out of curiosity?

Provided they are used as a supportive tool (not as “the unerring answer to everything”), and are checked by a human to ensure that no mistakes have been made, they can be incredibly useful, and can create excellent output!
 
Matt N said:
What exactly is the issue with the park using generative AI tools, out of curiosity?

Provided they are used as a supportive tool (not as “the unerring answer to everything”), and are checked by a human to ensure that no mistakes have been made, they can be incredibly useful, and can create excellent output!
Click to expand...
@Electric Elly, do you want to jump in? 😉
 
DistortAMG said:
Your best bet is to fight fire with fire and use one of the many free, AI powered, AI image detectors online.
Click to expand...
in my experience they often don't work, even o some obvious AI images they are like 50/50, I also think AT (if they have used it) have don addition photoshopping (adding at least differnt AI images such as the dad/child, crowd, towers, fireworks swirl)

IMO the dad and child looks real, the firework swirl I feel like I have seen before, although could be AI (although the GAL train is real, they have used it else where), the crowd could be AI or real (ard to tell too dark, although I feel like it is real for a few reasons) and the towers is again 50/50 (although I feel like it matches with the feel of the towers inside their map (not same model of course))

Towers inside the map: 1749574487536.png
 
AI is built into the Adobe suit by default now. It is pretty hard for anyone to stay away from it when it forms the core of industry standard image editing software.

This could be why it is referring to it being 50/50? As generally, you dont use AI to make the whole picture in pro image editing software, just to touch up and add bits to the image.
 
Last edited:
This is bit of a non-issue, the image is consistent with the graphic design Merlin/Towers have used for years. They’ve clearly layered up different images together and the crowd photo could be a stock image that happens to have been generated by AI.

If it is AI, it’s supplementary and a minor part of the image to the point most people won’t even notice. It’s been used meaningfully and with intent. It’s not the lazy AI-slop that we’ve seen from the likes of Drayton Manor and Phantasialand.
 
NuttySquirrel said:
@Electric Elly, do you want to jump in? 😉
Click to expand...
Gladly.

Ethically and environmentally, it's a mess.

First of all, a distinction must be drawn between AI and GenerativeAI. AI can be a fantastic thing in some cases, and has been around for a long time. GenAI is what I, and many other creatives and people, have a problem with.

How many hundreds, thousands, hell millions of artists have had their art and creations scraped and plagiarised for the sake of GenAI? It's effecting jobs worldwide for creatives- not just visual artists, but musicians, voice actors, editors, graphic designers, the list goes on. Why hire a human that can create something and pay them when "oh ChatGPT make me some slop with fireworks!" does the trick? It's not about speed (some artists are FAST), it's about cost cutting and not having to pay a human.

There is no regard to copyright in GenAI, especially imagery, which is something that OpenAI are lobbying for relaxations in the fair use policy so they can continue to train and scrape and ultimately profit off the backs of others, whether this is from large media companies or individuals who create for a living or hobby.

Yes, you can argue "but artists copy and draw from photos all the time!" and similar, but a human learns, a human observes and takes the nuance in what they are referencing and uses that as a skill to develop their own work and style, and most importantly gets a sense of enjoyment and fulfilment from creating. It has purpose. GenAI just generates patterns to best fit a prompt, and with high use of something like the "make me look Ghibli style oh computer" trend influencing and "inbreeding" its data, hence the p*ss yellow filter seen on a lot of 2D/cartoon style generations currently, and the same floppy plastic look it all has.

Morally as well, you need to look no further for an example than what was being done in excess on X/Twitter with Grok this weekend. Women were having their selfies over run with men requesting Grok to "make her eyes bigger and cross eyed, add glue on her face and a blush to her cheeks". Disgusting, and spread like fire over that hellsite. Just because a certain caliber of men cannot behave and think that all women are things to be objectified. This leads into deep fakes, and how easy it is for anyone to feed one of these LLMs with images of people- it could be someone famous, a colleague, someone they don't like- and generate images of them in compromising situations, acts, and more. It's done with voices as well, and is already happening and gets harder and harder to spot in some cases.

Environmentally as well, the data centres needed for GenAI models are massive. Microsoft and Amazon for example have both bought up land in drought prone areas worldwide to build their data centres, because it's cheap. Farmers and locals in Zaragoza express concern as despite the EU and Spanish government praising Big Tech for setting up their data centres, where does the millions of gallons of water come from? Already strained water reserves. Yes, these data centres aren't a new thing, but the rate and size they're being built for the purpose of GenAI is much greater than previously seen. It's not sustainable in multiple meanings of the word. It's estimated that every 20 to 50 prompts you give to ChatGPT can consume about 500 milliliters of water, which adds up very quickly with how integrated AI is in our searches like Google as default that can't be turned off in some cases.

I'm going off topic, but there are many more negative effects of GenAI, such as those seen in education and the effect on children's problem solving skills.

Please, also draw the distinction between a company using genAI imagery and a single person messing about once. I still disagree on a personal level with the use of these LLMs on a whole, especially with repeated use and using for the sake of profit (notably the surge of booths in Artist Alleys at conventions that pop up, which with a lot of conventions is a very competitive scene to get a table for, with every AI slop booth depriving an artist day/s of profit as there are some who make their living from conventions), but when a company is doing it, they're clearly not seeing the value in artists and designers and most likely, have found a way to shave off a person's wage to give to the CEO as a bonus. It's lazy, sloppy, and disingenuous.

Especially within the theme park industry, that we all love and support, it's an insult to the creative spirit, vision, and talent that is put into creating the experiences that we all cherish.

GenAI needs regulation, and it needs it last year. Perhaps with regulation and proper use, it can be used as the tool that people insist is its use, and not the abuse machine it actually is. Accessibility is not an argument, creativity and art in its many forms has always been accessible.

I can't apologise for the length of this, I'm tired of GenAI. Create, make a bad photoshop, scribble on a piece of paper, it's more fulfilling.
 
Some of the scariest things with GenAI is the legal stuff. Could a video be forged easier? Of course, it was possible before, but then someone who find out you spent hundreds commissioning either actors or that you racked up a huge VFX bill.. so nice try but that isn’t happening.
AI though? The only trace being of the account you used to sign in? That’s tricky. Seeing is not believing, it hasn’t been that way forever, we as humans have a narrow range of senses that leave us more unaware than not. But it’s going to get worse, legally anyway. However the worst has got it be the pervy stuff. Hearsay from news tells me there’s already ways to stick celebrities faces onto things… truly disgusting.

It’s like a Pandora’s box, with most of the things being coming out of this being pretty awful, and I suppose our hope equivalent shall have to be Will Smith eating spaghetti. But if we are to read that tale alternatively, where hope is too an evil, then I think our analogy could still function for a lot of people. Image and video gen? Really! These programs take time and effort to make, over that time who didn’t think this was a bad idea?

Back on topic a bit, did anyone find fireworks last year a peculiar show? It seemed like big tech corporation Phalanx showboats a lot and then saves the audience from aliens with big shooty shooty bang bang guns that go ratatatat. Aliens from space. Space…. Tech company… I honestly had this weird sense of discomfort watching a video of it? I mean really, it seemed a bit out of touch those 9 months ago, even outer of touchier now! Why instead of the Phalanx - who seem to be evil a lot - destroy these alien as we haven’t heard of before and then we all clap, the Nemesis monster eats them. Y’know, Kaiju stuff? I Take it that might be a bit scary for the kids, but I’d take it over the big fake corporation being really cool and awesome and saving everyone.

Am- am I going a bit bonkers? Please tell me I’ll have no shame!
 
Electric Elly said:
Gladly.

Ethically and environmentally, it's a mess.

First of all, a distinction must be drawn between AI and GenerativeAI. AI can be a fantastic thing in some cases, and has been around for a long time. GenAI is what I, and many other creatives and people, have a problem with.

How many hundreds, thousands, hell millions of artists have had their art and creations scraped and plagiarised for the sake of GenAI? It's effecting jobs worldwide for creatives- not just visual artists, but musicians, voice actors, editors, graphic designers, the list goes on. Why hire a human that can create something and pay them when "oh ChatGPT make me some slop with fireworks!" does the trick? It's not about speed (some artists are FAST), it's about cost cutting and not having to pay a human.

There is no regard to copyright in GenAI, especially imagery, which is something that OpenAI are lobbying for relaxations in the fair use policy so they can continue to train and scrape and ultimately profit off the backs of others, whether this is from large media companies or individuals who create for a living or hobby.

Yes, you can argue "but artists copy and draw from photos all the time!" and similar, but a human learns, a human observes and takes the nuance in what they are referencing and uses that as a skill to develop their own work and style, and most importantly gets a sense of enjoyment and fulfilment from creating. It has purpose. GenAI just generates patterns to best fit a prompt, and with high use of something like the "make me look Ghibli style oh computer" trend influencing and "inbreeding" its data, hence the p*ss yellow filter seen on a lot of 2D/cartoon style generations currently, and the same floppy plastic look it all has.

Morally as well, you need to look no further for an example than what was being done in excess on X/Twitter with Grok this weekend. Women were having their selfies over run with men requesting Grok to "make her eyes bigger and cross eyed, add glue on her face and a blush to her cheeks". Disgusting, and spread like fire over that hellsite. Just because a certain caliber of men cannot behave and think that all women are things to be objectified. This leads into deep fakes, and how easy it is for anyone to feed one of these LLMs with images of people- it could be someone famous, a colleague, someone they don't like- and generate images of them in compromising situations, acts, and more. It's done with voices as well, and is already happening and gets harder and harder to spot in some cases.

Environmentally as well, the data centres needed for GenAI models are massive. Microsoft and Amazon for example have both bought up land in drought prone areas worldwide to build their data centres, because it's cheap. Farmers and locals in Zaragoza express concern as despite the EU and Spanish government praising Big Tech for setting up their data centres, where does the millions of gallons of water come from? Already strained water reserves. Yes, these data centres aren't a new thing, but the rate and size they're being built for the purpose of GenAI is much greater than previously seen. It's not sustainable in multiple meanings of the word. It's estimated that every 20 to 50 prompts you give to ChatGPT can consume about 500 milliliters of water, which adds up very quickly with how integrated AI is in our searches like Google as default that can't be turned off in some cases.

I'm going off topic, but there are many more negative effects of GenAI, such as those seen in education and the effect on children's problem solving skills.

Please, also draw the distinction between a company using genAI imagery and a single person messing about once. I still disagree on a personal level with the use of these LLMs on a whole, especially with repeated use and using for the sake of profit (notably the surge of booths in Artist Alleys at conventions that pop up, which with a lot of conventions is a very competitive scene to get a table for, with every AI slop booth depriving an artist day/s of profit as there are some who make their living from conventions), but when a company is doing it, they're clearly not seeing the value in artists and designers and most likely, have found a way to shave off a person's wage to give to the CEO as a bonus. It's lazy, sloppy, and disingenuous.

Especially within the theme park industry, that we all love and support, it's an insult to the creative spirit, vision, and talent that is put into creating the experiences that we all cherish.

GenAI needs regulation, and it needs it last year. Perhaps with regulation and proper use, it can be used as the tool that people insist is its use, and not the abuse machine it actually is. Accessibility is not an argument, creativity and art in its many forms has always been accessible.

I can't apologise for the length of this, I'm tired of GenAI. Create, make a bad photoshop, scribble on a piece of paper, it's more fulfilling.
Click to expand...
Your argument is compelling and passionate and is one way of looking at things. The moral arguments and environmental impact concerns you highlight are present in the "old" way of doing things though. I appreciate that I'm about to write a load of waffle that you already know, but this post is for everyone else who reads this thread and forum too.

Let's view this through the lens of how this particular poster would have been created before the dawn of AI, just a few years ago. In particular we're focussing on the crowd shot.

It's incredibly unlikely that Alton Towers would have commissioned a photographer to take a specific photo of a crowd for promotional use on their poster. It's likely that they would have either used photos from previous events, or more than likely relied on a stock image.

In the case of a stock image, a photographer will have taken a photo at some point in time and usually sold the licence to reuse that photo to a stock image library. They could have also uploaded the image to a free stock image service, which usually have conditions around commercial usage, but there are many which don't. Adobe run their own stock image service, but the largest paid for is Getty. Getty as in Getty Oil, which at its peak were one of the largest oil conglomerates in the world. Not a bastion of green, clean and efficient energy.

The servers which Getty, Adobe and other services use to host stock images take up vast amounts of space and energy. As they are constantly storing something in perpetuity, and their collections only get larger, so do their energy and space concerns. There is a burst of energy consumption when something gets created by GenAI, but it doesn't have to be stored in perpetuity. A rather crude analogy would be do you always want to have your house heated to 18˚C, and use a little bit of gas throughout the day to keep it toasty, or do you only put it on when you need it and use a moderate amount of gas in a short period? The latter seems worse, but uses less gas in the long run.

If Alton Towers obtained their crowd image using Getty (and let's presume they're going for a royalty free licence), the artist would have been paid a fraction of Getty credits for the use. When the credits hit a certain amount, $50, $100, etc, the artist can check out those credits for cash. For royalty-free content, the rate is typically 20% for still images and 25% for video clips. The artist is arguably getting paid, but it's a small amount with the lion's share going to Getty. There's the additional kicker that Getty have also licensed their entire library for GenerativeAI scraping.

The photograph would have been taken on a professional camera. 5 years ago this would typically have been either a Canon or Nikon. The camera body would have been manufactured in China, or Taiwan. The factory workers will not have been paid a fair wage by our standards and, depending on the region the body was assembled in, could have been subject to human rights abuses.

High end cameras are made using rare earth materials and plastics. It is incredibly energy intensive to manufacture a digital camera and its lens. This camera will only serve one person at a time, unless it's in a hire house, but it doesn't have mass simultaneous use. The same is true for any accessories, batteries and digital storage which are required.

The photographer will have then categorised and touched up the photo on a computer. It's likely that the device will have been manufactured by Apple, who also have questionable morals when it comes to the rights of factory workers in their manufacturing centres. If we're focussing on founders and CEOs, you'll be in warm company with Steve Jobs. Largely considered one of the most toxic company leaders of modern times. This is before we turn our attention to privacy focussed Apple, who will happily protect your privacy rights... unless you're in China. This computer will have also been made out of rare materials and energy intensive to manufacture.

Your camera body and your computer will both be obsolete within 5 years, you'll have to buy a new one and the process repeats.

The depiction you've described this weekend with Grok and X is horrific and upsetting, but it sadly isn't new. It's been possible to do these things within standard Photoshop for years, there are many places on the internet dedicated to it. The main difference is the ease at which more people can do these things and how fast they can, without training. There have been decades worth of arguments about the ethical concerns about using Photoshop, which I'm not going to repeat here, but this standard industry tool was once considered the work of the devil too.

In the specific case of this particular poster, and the image of the crowd shot, there are benefits and downsides to using GenAI, or stock image photography. Both have their complex concerns issues and nuances, it's not clear cut. I appreciate that you care about this very deeply though, and it has a personal impact, but we can always look at things in more than one way.
 
I'm unsure if Towers have used AI to create that promo image, but it would certainly be a pity if they had. But also not a surprise with how Merlin operate these days.

There have been some good points made on the use of GenAI, but if we can now stick to Alton Towers discussion then that would be great!

Good news that Fireworks is back to 3 nights, they should never have cut it to 2 in the first place. I still have no plans to renew my MAP once it expires later this year though, needs the likes of Oktoberfest to return for me to feel like I am going to get value.
 
You must log in or register to reply here.
Top