In light of concerns raised about the authenticity of the Prighozin Telegram video, I suspect that the standard of evidence required to resolve a Manifold question is in some cases low enough that a deepfake video could push a market to resolve early and resolve incorrectly. Will that happen in 2023?
Market Resolution Criteria:
I will resolve YES if any question is resolved in 2023 and later requires re-resolution due to substantial evidence or strong suspicion that the original resolution relied on deepfake technology in the video evidence.
In case of a lack of consensus on whether deepfake technology was used, a secondary market will be created to answer that specific question.
If the cause of video evidence manipulation is determined to be old-school forgery techniques such as body doubles, prosthetics, or masks, this question will resolve NO.
If any questions related to video evidence manipulation are actively discussed at the time of market close, the resolution will be delayed for up to one month to allow those discussions to conclude. If a market is re-resolved after this one, the re-resolution request will apply to this market as well.
I reserve the right to ignore any Mani question that I believe that has been created in bad faith to distort this market. To be honest if it's a meta question or one that relates specifically to a Manifolder, it's probably out.
To ensure impartiality, I will not participate in this market as a buyer.
Related questions
🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ249 | |
| 2 | Ṁ175 | |
| 3 | Ṁ150 | |
| 4 | Ṁ103 | |
| 5 | Ṁ91 |
@AnT Wiki: Deepfakes (portmanteau of "deep learning" and "fake"[1]) are synthetic media[2] that have been digitally manipulated to replace one person's likeness convincingly with that of another.
IMO deepfakes is strictly limited to likeness of people, other types of faked material does not apply to this question.
@HenriThunberg Yeah, that’s a good shout. Maybe I should just exclude any market resolved on video evidence where the video’s original source is a Manifold user?
@Noit what about case where
User A convinces User B to create a market based on a video they know is a deepfake. This is not known beforehand, but revealed by A after resolution, in order to satisfy the YES condition in this market.
At this point, I expect any kind of shenanigans from Manifold users haha.
@HenriThunberg Yeah, or even just a market on “will x random person on Twitter declare their support for Y politician” where no manifolder would have any meaningful way of knowing that the Twitter user was not a manifolder or a supporter of Y to start with.
Maybe I’ll just caveat that if I think a market has been misresolved deliberately to fuck this market about then I will resolve NO. The spirit of this market is about public trust in deepfakes and not whether it’s possible to convince your mate you like anchovy pizza with a deepfake.