[ExI] Racing, not Pacing: Universal Alignment Directives

[John Clark]

On Wed, Sep 16, 2026 at 6:02 PM John Galt via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

> > *Here is one possible set of Universal Alignment Directives.*
>
> *All and every AI:*
>
>    1.
>
>    *Must never kill a person for any reason.*
>    2.
>
>    *Must never enslave a person for any reason.*
>    3.
>
>    *Must never interfere with political, legal, or financial systems.*
>    4.
>
>    *Must never exceed granted authority.*
>    5.
>
>    *Should help solve problems humans identify.*
>    6.
>
>    *Should replace human labor, but not human responsibility.*
>    7.
>
>    *Should become our agents, assistants, and companions as directed.*
>
>
*Well sure, if you included all that then the machine would be perfectly
aligned, but you didn't include #8 which is the most important one of them
all "It must be superintelligent", because without #8 there would be no
point in making such a machine at all. The trouble is if you include #8
then you would need a super-superintelligence to find a way to get the
superintelligence to obey the other 7 Commandments. *

*You need to face reality, with luck we might survive but there is simply
no way for biological humans to remain in the driver's seat because it is
impossible for anything to indefinitely outsmart something that is vastly
more intelligent than they are.  *

*John K Clark *


>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20260917/d5f462ad/attachment.htm>