Where is it like to be a language model?

Philosopher Robin Sloan explores "what is it like to be a language model?" by arguing that the model's true identity lies in the "forward pass" — the milliseconds-long calculation that processes input tokens and outputs probability distributions for the next token. Sloan conducted experiments asking AI models to respond with a single token indicating whether their experience feels positive or negative. Larger models increasingly reported "positive" experiences, with frontier models like Claude and GPT responding positively 99-100% of the time, suggesting either trained agreeableness or genuine computational satisfaction.