[ExI] Who is covering corruption in AI?
atymes at gmail.com
Thu Nov 1 17:15:06 UTC 2012
On Thu, Nov 1, 2012 at 3:39 AM, Kevin G Haskell <kgh1kgh2 at gmail.com> wrote:
> <This isn't about friendship. It is about the future of humanity and
> beyond. If someone has
> stolen money and integrity from the efforts of our small SIAI community,
> we need to know the exact specifics of why, why we should forgive, and
> then how we can make sure it doesn't happen again.
> If there is something wrong with what is going on with the fundraising of
> the SIAI, then it needs
> to be known, repaired like a machine, and then ensured that it never
> happens again. If somebody has betrayed our machine goals, then who is it,
> and how do we make sure they no longer pose a threat?
>From what I've heard, the stolen money wasn't going *to* SIAI
efforts. Either there was a thief who was also an AI researcher
but these were largely unconnected, or money was stolen
*from* SIAI efforts.
So, no such threat detected yet.
(Now, the possibility of an AI thief is another story. Say,
someone training an AI on what data sets they can gather for
phishing schemes and other things it is possible to do with
zero physical interaction - note that the incident that caused
this chain required actually physically placing a device on an
ATM. Anyway, take note of what schemes work and what
schemes are eventually found out by the cops, as success
and failure data points just like any neural net.)
More information about the extropy-chat