The objectives of AGI are excellent.  Conceptually the AGI model can be
regarded as an extension of AI.

However, pragmatically everything that works today  from chess to AlphaGo,
heuristic search, ANN combined with statistics (e.g. machine learning) *is
“narrow”  AI not AGI*.



To my knowledge, in the last 15 years none of the new "techniques" has
shown any AGI breakthrough. If you feel that we are wrong just provide a
single example of an AGI algorithm that works and it is not "narrow".
Science fiction writers everywhere they pretend to enlighten the crowd .
Since 70s they envisioned building autonomous systems, and hollywoodian
scenarios where highly intelligent computers  took control  e.g.
https://en.wikipedia.org/wiki/WarGames . Later the deadline for such
digital computers was extended to   2000,   2010 and 2020......



 Today we know that such apocalyptic expectations are "fake ", PR stunts.
Andrew Ng posted recently on twitter  "Let's cut out *the AGI nonsense*".
Without real progress no one will believe or invest a single dollar  in
"AGI algorithms". So the future of AGI on digital computers is bleak.



*The AGI level will not be reached solely by developing algorithms on
digital computers*. Only a few understood ten years ago the real
problem.  That's
why Markram's initiative and all other  brain projects are today in
disarray,
https://www.scientificamerican.com/article/why-the-human-brain-project-went-wrong-and-how-to-fix-it/
, see
also Modha's DARPA SyNAPSE  and many other programs. There is no “digital
brain” and singularity is nowhere near. Everything was a  PR stunt. Have we
learn anything from such failures? No, not enough to make a single step
further.

And even worst,   after over 15 years of PR we cannot see a glimpse of
AGI on digital computers.

 *Why?*

 *The difference between computing by interaction and algorithmic
computations run on digital computers is huge (see neuroelectrodynamics,
NED).* "Physical boundaries" are highly important since there is a level
of interaction required for any computing system to be "AGI compatible”.
The brain computations by interaction (or vibration if you like)
provide the best example of the new physics of computing that’s highly
“general"  not “digital”. Such systems have to be built first if someone
really likes to push the field of computation forward and at this point in
time probably China has the resources to make this step.



The new space of computing by interaction will be way different than the “
*Jurassic* World of algorithms” performed on digital computers. This new
field needs investors and I hope this brief introduction will help Colin,
Ben and maybe others to secure the required funding.



Dorian


------------------------------------------
Artificial General Intelligence List: AGI
Permalink: 
https://agi.topicbox.com/groups/agi/Ta6fce6a7b640886a-M23cc8f41369e6684de61f478
Delivery options: https://agi.topicbox.com/groups/agi/subscription

Reply via email to