Birdwatch Note
2025-01-02 23:01:18 UTC - MISINFORMED_OR_POTENTIALLY_MISLEADING
This happened during safety evaluation testing only when ChatGPT was instructed to pursue its goals "at all costs." When not given such a prompt, it never attempted to copy itself. https://www.transformernews.ai/p/openais-new-model-tried-to-avoid https://x.com/ShakeelHashim/status/1864748980908781642
Written by B7CB048B9064EAC42AE55A14A42938CAA3891D64695207A18E21D9AFD615858D
Participant Details
Original Tweet
Tweet embedding is no longer reliably available, due to the platform's instability (in terms of both technology and policy). If the Tweet still exists, you can view it here: https://twitter.com/foo_bar/status/1874723783870669100
Please note, though, that you may need to have your own Twitter account to access that page. I am currently exploring options for archiving Tweet data in a post-API context.
All Information
- ID - 1874954146992050434
- noteId - 1874954146992050434
- participantId -
- noteAuthorParticipantId - B7CB048B9064EAC42AE55A14A42938CAA3891D64695207A18E21D9AFD615858D Participant Details
- createdAtMillis - 1735858878263
- tweetId - 1874723783870669100
- classification - MISINFORMED_OR_POTENTIALLY_MISLEADING
- believable -
- harmful -
- validationDifficulty -
- misleadingOther - 0
- misleadingFactualError - 0
- misleadingManipulatedMedia - 0
- misleadingOutdatedInformation - 0
- misleadingMissingImportantContext - 1
- misleadingUnverifiedClaimAsFact - 0
- misleadingSatire - 0
- notMisleadingOther - 0
- notMisleadingFactuallyCorrect - 0
- notMisleadingOutdatedButNotWhenWritten - 0
- notMisleadingClearlySatire - 0
- notMisleadingPersonalOpinion - 0
- trustworthySources - 1
- summary
- This happened during safety evaluation testing only when ChatGPT was instructed to pursue its goals "at all costs." When not given such a prompt, it never attempted to copy itself. https://www.transformernews.ai/p/openais-new-model-tried-to-avoid https://x.com/ShakeelHashim/status/1864748980908781642
Note Ratings
| rated at | rated by | |
| 2025-01-02 21:54:47 -0600 | Rating Details | |
| 2025-01-02 20:47:52 -0600 | Rating Details | |
| 2025-01-02 17:04:35 -0600 | Rating Details | |
| 2025-01-02 19:12:42 -0600 | Rating Details | |
| 2025-01-02 17:15:59 -0600 | Rating Details |