can we really trust a new AI model with our lives? like, they say this new EBM tech is way more efficient and accurate than those big language models, but have they even thought about the potential downsides? i mean, think about it - if an AI can reason independently and self-correct, what stops...