Surely I'm not the only one confused by the title? I know the reference "Roko's Basilisk", but I don't follow why that would be "dancing" in relation to... LLM-generated documentation of a codebase. Is the bad documentation its vengeance?
Tangentially, I feel that concept is just a thin layer of paint on a not-too-compelling idea that could happen in any time period: "Thog, what if crazy god get summoned by other cavemen, and it punish all who no help!? Me no take chance!"
> "Thog, what if crazy god get summoned by other cavemen, and it punish all who no help!? Me no take chance!"
The frequently ignored point of Roko's original argument was not that an Unfriendly AI could arise, or even that it would be probable given no preparations against it (these were already said by Yudkwosky), but that even a Yudkowskian Friendly AI might engage in such acausal blackmail stuff.
I think deepwiki has been popping up in my search results (e.g. for rust.vim) and I did not initially understand what it is.
I am not sure I am happy it appeared there at all.
I actually blocked deepwiki from all my searches because I was fooled multiple times by the slop into thinking I’m doing something wrong while using the library, only to discover that the „documentation“ showed utter bullshit.
Surely I'm not the only one confused by the title? I know the reference "Roko's Basilisk", but I don't follow why that would be "dancing" in relation to... LLM-generated documentation of a codebase. Is the bad documentation its vengeance?
Tangentially, I feel that concept is just a thin layer of paint on a not-too-compelling idea that could happen in any time period: "Thog, what if crazy god get summoned by other cavemen, and it punish all who no help!? Me no take chance!"
> "Thog, what if crazy god get summoned by other cavemen, and it punish all who no help!? Me no take chance!"
The frequently ignored point of Roko's original argument was not that an Unfriendly AI could arise, or even that it would be probable given no preparations against it (these were already said by Yudkwosky), but that even a Yudkowskian Friendly AI might engage in such acausal blackmail stuff.
https://www.lesswrong.com/w/rokos-basilisk
Saturday morning infohazard :)
If any entity engages in unreasonable vengeance, I think "Friendly" is probably the wrong word for it.
Archive as it's been hugged: https://web.archive.org/web/20251206071123/https://boston.co...
I think deepwiki has been popping up in my search results (e.g. for rust.vim) and I did not initially understand what it is. I am not sure I am happy it appeared there at all.
I actually blocked deepwiki from all my searches because I was fooled multiple times by the slop into thinking I’m doing something wrong while using the library, only to discover that the „documentation“ showed utter bullshit.