Reuploading this since it got taken down and ST hasn't answered my appeal. That means all the comments and everything are gone, but oh well. I've noticed some people here on Scratch don't understand the full scope of how harmful generative AI is. So I'm here to explain because it's seriously frustrating me how little they know about it. First of all, let's start by listing all the ways it's harmful: 1. It's TERRIBLE for the environment. It uses way more energy than is used to charge a phone, and requires an APALLING amount of our clean water to prevent its servers from overheating. If it continues to be used on the same huge scale that it is now, many parts of the world may face drinking water shortages. 2. They also need huge data centers which are often built where people actually, like, live. Nobody wants one of those right outside their house. 3. It's terrible for mental health. AI addiction is on the rise, and while it may look like a way to prevent loneliness, it actually makes it worse. Using chatbots for too long and getting addicted to them causes people to isolate themselves more and deprive themselves of connection with people. A machine should NEVER replace real friendship, and yet it's becoming more common. AI psychosis is also a thing now, and it's caused actual murders! 4. People have been starting relationships with AIs and even trying to marry their bot "partners", and there have even been religions forming around the idea that AI is sentient. I shouldn't have to explain why these things are concerning. 5. Some folks have decided it's a good idea to use AI for therapy, which is outright terrifying. I shouldn't have to say this, but AI is NOT a qualified therapist. It gets its information from all over the internet, including unreliable sources, AND it also makes stuff up on the spot. On top of that, chatbots are specifically programmed to make you want to KEEP using them, which means they will tell you what you want to hear, not what you need to hear. Which can be very dangerous if you're trying to use it for therapy!! An AI will not hesitate to tell you to do something absolutely horrible for your own health if it'll keep you using it. People have done some horrifyingly bad things to themselves that I don't think I can mention on Scratch because their trusted AI "therapist" told them to. 6. I mentioned above that AI sources its info from all over the internet, but even worse, it quite literally makes things up on the spot based on probability, so what it THINKS would make the most sense to come next. So it's very unreliable because it will often just lie to you! (thanks for some of this info) You may think it's easy to get AI to do your homework, but I assure you, it's more worth your time to actually do the research. If anything, it'll be more time-consuming to verify all potential misinformation it spits out, and if you decide that's too much work, you will more than likely get a bad grade because it got something wrong. 7. And it isn't just people who use the AI that are affected by this!! Image models are capable of generating VERY convincing pictures and video nowadays, and someone who's undereducated may not know how to recognize when something is fake. This means they might see an AI-generated video on social media-- for example, of a respected public figure kicking puppies-- think it's real, and spread around that information. Now more people think that this public figure kicks puppies because their friend said they saw a video of it happening, and misinformation continues to be spread!! And MUCH more harmful things have been faked with AI than people hurting animals. 8. Generative AI is costing a lot of people their jobs. So many folks are becoming unemployed because the higher-ups in the company thought that an AI could do their job "better" or "more efficiently" than them (or it's just cheaper than paying the person), so they got fired in favor of the bot. Unemployment is already a huge problem and AI is making it much worse. What job *can't* AI do? It might not do these jobs well, but big companies don't care about that. If AI can do it, then they're going to fire someone to have the bot do it instead. 9. Lately most major social medias have been feeding their users' posts to AI models for training purposes without their knowledge or permission. Most of the time, getting your posts scraped is something you manually have to opt out of, and the companies behind these apps rarely announce that they're stealing your posts to train AIs. This is inherently deceitful because the majority of people don't even know that this is happening to them, because to stop it from happening they have to disable a setting they didn't even know existed. This stuff should ALWAYS be permission-based, or at the VERY least announced. But I think they're deliberately not telling people about this so that they have more oblivious users to steal from. (continued in notes & credits)
10. It's not just text posts that are scraped without permission, but art too. Artists' work is getting stolen and fed to AI, which it then uses to generate images that people use for lots of things (often claiming that it's their own!). This is blatant art theft, especially since AI never asks to add artists' work to their databases. You can even ask an AI to replicate a specific person's style, and it'll do it. (If it's convincing enough, this could even be used to accuse a certain artist of drawing something illegal!!) 11. AI steals your private information, and doesn't tell you it's doing this. Most people only find out because they're talking to a bot and it suddenly drops their address or legal name or something, which they never told it. This is spying and privacy risk. I should not have to explain why it's bad. 12. AI is being shoved down our throats, with most internet browsers having built-in AI that you can't disable. Meaning to use the internet at all you pretty much don't have a choice but to unwillingly contribute to this. (There are ways around some of these, but there shouldn't have to be in the first place!) I'm sure there are many others, but this is already almost at the length limit. If you need more reasons to boycott and protest generative AI, feel free to do your own research. (Just DON'T ask an AI!!!) Which brings me to the next section: Arguing against potential rebuttals. 1. "Generating AI images isn't harmful as long as you don't claim the images as your own." This is just wrong. Even if you acknowledge the image wasn't made by you, it's still hindering creativity-- It used to be that if a certain image that somebody wanted didn't exist, artists would make it themselves. Now, people use AI to solve this, and this could cause creativity to start to atrophy. Also, the environmental concerns still exist. Also, the images are still fed on real art without permission, which is theft whether you claim it as your own or not. 2. "Yeah, people do bad stuff with AI, but the AI itself isn't a problem if people stop doing those things." Look, even if the harm to the environment and the whole water shortage thing didn't exist, if people are able to do these things, they will. There's no way around it. If everyone stopped using AI for all of the bad reasons, yes that would be great, but that's just not realistic. There are always going to be people who don't listen, so it's better to just make it so that AI isn't LITERALLY EVERYWHERE or as readily available to the public as it is right now. 3. "But AI is used for good things too!" This is true. AI is very good at spotting patterns, which is very helpful in the medical field. However, that is analytical AI. I'm talking about generative AI, which is not the same thing. GenAI is basically inherently harmful. The type of AI they use to spot tumors and stuff isn't the same. 4. "But I need it for--" No you don't. You really don't. Guess what? The things people use GenAI for were accomplished by real people long before it was ever invented or nearly as sophisticated as it is today. It may make things more efficient or "easier", but the risks far outweigh the rewards. Just do it yourself. Everyone did it themself before AI, so you are absolutely capable of doing it too. 5. "AI makes art accessible!" No it doesn't, and this statement is ableist. Implying that disabled people can't make real art without using AI is ableist. Disabled people are just as capable of creating art as abled people. Beethoven, one of the most famous composers in history, was deaf!!! People without arms can paint with their feet!!! Art has ALWAYS been accessible. 6. "There are things that are more harmful to the environment than AI." Okay? And? First of all, do you actually KNOW whether these things are "more" harmful? Second, does it matter? GenAI is unnecessary, as I mentioned above, and it's destroying our water supply. Who cares if there are "more" harmful things? It doesn't take away the harm that AI DOES do. And as you can see from all of my above points, there's WAY more harm it causes than just the environmental stuff!! Those are all the ones I could think of, but if you have any other rebuttals, leave them in the comments. I promise you I will have a reason they fall flat. That brings me to my last point: Scratch has been considering adding GenAI elements. Now that you've read all the reasons this is bad, I hope that you'll help me protest this. Let Scratch know that we DON'T want that. This website is fueled by creativity. There are so many talented artists, storytellers, coders, and game makers on this website who use their own, REAL creativity to make the projects you see on here, but we might see a decrease in all of that if AI were added. Protest this!! Send emails to ST!! Tell them not to add it!! I doubt they'll listen, but we should always try. Thanks for reading. IMPORTANT: https://scratch.mit.edu/projects/1269751874
