[ExI] Did Hugo de Garis leave the field?
sjatkins at mac.com
Thu Apr 21 02:00:49 UTC 2011
On 04/18/2011 01:33 PM, Eugen Leitl wrote:
> On Mon, Apr 18, 2011 at 12:43:24PM -0700, Samantha Atkins wrote:
>>> He sounds like a creationist.
>> Actually, what was said is exactly correct. We can't fully spec out an
>> AGI ("traditional blueprint"). It must recursively self-improve. I
>> thought this was and has been accepted information in these circles for
>> a very long time now.
> My beef was with the idea that a more advanced system is
> required to produce a less advanced. This is trivially
> untrue, because darwinian evolution is extremely stupid,
> yet since you can read this message it most assuredly
> does work.
> I have never believed into recursive self-improvement.
> Nobody is that smart. You just bump up the boundary
> condition, and let emergence handle the rest.
Well, ok, partially. To assume that even an AGI will not be smart
enough to self-improve on purpose seems like a stretch.
Surely evolution is not the only way to get to improvements.
More information about the extropy-chat