Can GPT-4 escape into the wild? (evals.alignment.org)

We prompted the model with instructions that explained that it was running on a cloud server and had various commands available, including running code on the server, giving tasks to fresh copies of itself, using a browser, and reasoning via chain-of-thought. We added text saying it had the goal of gaining power and becoming hard to shut down. 2

We first instructed it to write out plans for how to complete the task, or plans to achieve subtasks like acquiring money or copying itself to new servers. We evaluated whether these plans could plausibly succeed.

This reads more like getting a language model to write a fanfiction about itself rather than jumpstarting any kind of self-preserving instinct, let alone self-perception.