
https://www.bbc.com/news/articles/cd0el3r2nlko
https://www.theguardian.com/us-news/2024/dec/28/openai-whistleblower-suchir-balaji
https://www.firstpost.com/explainers/openai-whistleblower-suchir-balajis-suicide-fbi-13848584.html
Resolves when conclusive evidence comes out confirming an option.
Comment if you want me to add / edit options.
Because OpenAI doesn't gain anything from murdering a former employee, so supposing the evidence is inconsistent with suicide the motive is likely financial.
So I think it's highly unlikely OAI is involved directly, but maybe it may be related in that the perpetrator was invested in OAI and saw the lawsuit and data he had as worth millions.
More likely it's related to his future endeavors in creating a startup worth multimillions. Also can't rule out random murder given the high rate in San Fran
A lot of this is speculative and based on George Webb's reporting which comes to a conclusion that it's intentional and unrelated to OAI.
@ChinmayTheMathGuy SF actually has a lower homicide rate than the US national avergae, as of 2019
@BaryLevy Also, am I wrong in assuming that well-off people with good executive function being killed in their home represent a tiny fraction of the homicide rate? If it's a murder, then it's an extremely non-representative one anyway
See my compilation of evidence here: https://www.lesswrong.com/posts/DDtEnmGhNdJYpEfaG/joseph-miller-s-shortform?commentId=ihtKnKxxF2Z8ZKEGL
@JosephMiller0e84 i don't recommend reviewing crime scene footage, but the wad of hair, and blood smears from bathroom to hallway? I agree the rest of the evidence isn't meaningful, but I don't understand the update at a granularity of 2% from 98 to 96. Is there a track record to support this precision and confidence? Surely we can find out if he owned a gun, no?