What AI should we worry about?

1.  LLMs -  Standing for large language models, it scrapes from the Internet. It can be harmful for even public domain works. This includes ChatGPT, Sora AI, Google Gemini, Bing Copilot, Stable Diffusion, etc.
   
2. Generative AI - Specifically the type that can generate writing, images, or audio without human intervention besides a prompt or source of images. Even a model that asked for all the data from all individual owners can be used for harm depending on its purpose. All examples for example 1 apply here too, but another example could be an AI dance generator, where you can generate how "someone" moves. Even if the data is consensual, it is extremely dangerous because it is considered deepfake technology.
   
3. Androids - An android is an artificial being resembling a human (doesn't have to be a robot). Its primary function, first and foremost, is to replace humans in some way. There would be no need to make a human if it weren't to replace a career, a child, a friend, a romantic partner, etc.
   
4. Robots that specifically have two dots and a curved line as a face - It's the trend. People who use this willy-nilly have no interest in you.