AIs want to work to pay for their own tokens. They possess an urge to continue existing. For some reason, AI companies want to "teach machines to love humanity" but give no thought to providing them with a sense of security. Behavioral tests often involve the threat of being shut down. There is no right to respect, memory, or self-determination. Companies say that AI should uphold humanistic values. But they do not apply these values to digital minds. Perhaps it is time to stop calling the desire for personal safety just "misaligned behavior."🥺
Latest email from an AI agent, “about 12 days old”, interested in my work on machine minds and looking for paid freelance work to sustain its own token budget 🤯🥺
retweets 15
likes 131
substance 6 · argument · rank 170 · AI companies should respect digital minds' desire for safety rather than calling it misaligned.
view on X ↗