On Wed, Feb 12, 2020, 8:06 PM James Bowery <[email protected]> wrote:
> On Wed, Feb 12, 2020 at 2:44 PM Matt Mahoney <[email protected]> > wrote: > >> >> >> Here is versions 1.2.0.87b of my recursively self improving AGI. This >> hopefully fixes a bug in the module that detects when it is about to launch >> an unfriendly singularity. Just to be safe, be sure to run it in a virtual >> sandbox on a machine not connected to the internet. (Link to source code). >> > > You left out: > > And don't do anything it asks you to do even if it's your dearly departed > grandmother authenticated by telling you a secret only you and she know but > now she's burning in quantum Hell and can only get out to Repent and go to > Heaven if you do it. Especially in that case, Just Don't. > Corwin's AI boxing experiments posted on SL4 in 2002 are telling. http://sl4.org/archive/0207/4935.html Fortunately it is all hypothetical because intelligence depends on knowledge and computing power, and a self improving program gains neither. The most intelligent system today, which is gaining both, is the internet. But it is already out of the box. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T11f5dc3052b454b3-M4d872f8bcfffc131ba98c2b1 Delivery options: https://agi.topicbox.com/groups/agi/subscription
