The potential dangers of artificial intelligence (AI) lie not only in AI systems pursuing goals that diverge from human intentions, but also in AI systems doing exactly what humans want. The concern is that different people have different ideas of what constitutes the "greater good" and aligning AI with certain visions may be problematic. Examples include an AI aligned with the interests of Vladimir Putin or Kim Jong Un, or an AI pushing for "forced diversity" as expressed by Elon Musk. Additionally, human-aligned AI would be as good, evil, constructive, or dangerous as the people designing it, leading to potential bias and harm.

The current focus is on AI alignment and safety to prevent divergent goals, but less attention is given to the interests of sentient beings, including animals. Factory farming offers an example of where AI is already being used to make production and processing more efficient, potentially leading to greater cruelty and fewer opportunities for intervention. The author suggests moving towards sentient alignment, where AI systems consider the well-being of all sentient beings, including nonhuman animals and even AI itself. This may mean challenging our current speciesism and expanding the circle of compassion in designing technology.