AI is not intelligent! At its core, Generative AI is a very powerful guessing machine that is good at working out what letters, words or visuals should come next in a sequence. This means it can present a simulation of thinking or creation. But this is inevitably limited by the data that the AI models are trained on. Do we really want to outsource the work that our brain does to understand and interact with the world? We say no.
AI is being forced on us. It is becoming harder to avoid - we believe that we should have a choice in whether we use it. We know that billionaires invested in AI and their companies are deeply embedded within repressive governments and the defence and security industries. Embedding AI in all levels of everyday life is giving them more control. We should have a choice in whether we allow this.
Theft is at the heart of all the mainstream large language models. They have been trained on data that's been taken from artists without their consent, and is now used to mimic their work. The copyright situation for AI-generated content is entirely unclear - it's being contested whether it can be copyrighted at all, and there is discussion on whether AI companies can claim copyright or at least partial ownership of ideas generated with their tools.
AI cannot be creative in a true sense. It has no intentionality (it does not know what it is trying to do and say), no interiority or self-awareness, and no frame of reference outside of the models that it is trained on. AI can create a synthetic impression of creative work, in many cases very effectively. But it has no imagination. And we feel that to allow AI output to be called creativity or ‘art’ hugely diminishes creative practice.
AI and ‘machine’ learning often relies heavily on invisible human labour, from data entry and labelling to content moderation. These jobs are often held by a precarious labour force in the Global South, and it is their exploitation that is powering the growth of AI industry.
The environmental impact of the AI industry is manifold. Seeing this technology being deployed at all levels of society during a climate crisis should make us pause. The benefits from a few use-cases are negated by the enormity of the environmental consumption of all the others.
AI powers the social media slop machine. Ranging from mediocre to offensive or dangerous, that content is exploding with the AI boom and threatens to overwhelm the Internet.
Most AI output is mediocre. We see this every day. The marketing copy that makes no sense, the news story that has no structure, the email summary that misses the key point. It’s lazy, and frankly insulting, and we don’t see why we should have to put up with it.
AI is not safe. It is often heralded as a miracle pill for the lack of affordable mental health provisions, and suggested to teenagers as a tool to improve their wellbeing. However, there have been multiple, documented cases of AI psychosis, and young people are increasingly withdrawing from socialisation with their peers because the sycophantic chat bots are easier to chat to. Vulnerable populations are more at risk of coming into harm from AI, yet they are the ones that are often cited as needing it the most.
Students are asked to use AI for their work but also to verify its output. Workers are coerced into using it, yet need to fix its mistakes. Efficiency savings resulting from AI have been found to be fictitious. And even if this wasn't the case, your boss would just find more for you to do anyway. It's a deskilling process that benefits no-one, except a narrow class of executives who can use it as a surveillance tool. We believe in the right to say no, and to let our humanity shine through.