[x3d-public] AGI anyone? Hidden in plain sight

John Carlson yottzumm at gmail.com
Sun Dec 11 19:04:34 PST 2022


What i envision (barely visually), is that back in the first half of the
20th century, people were trying to build functions with systems as
parameters, and stuff like the halting problem popped out, you couldn’t
determine whether the function would halt or not.

So Turing made sure that all inputs were bits, and things worked well, or
at least we could reason easily about machines and programs.

Nowdays, we have new inputs which haven’t been widely covered in a Turing
Machine:  qubits, pixels and voxels (or have they?).  We also have better
ways to describe systems: modeling languages, specification languages etc.

We have good ways of reasoning about pixels with Convolution Neural
Networks and voxels with MineCraft and we also have quantum hardware.  Can
we fully reason about systems yet?   We have metamodeling, object
constraint languages, maybe others.   We have MTbD and xslt to transform
models.

Do we have Deep Systems yet?  Are we still clueless about reasoning about
them even though we can build them?  Or do deep systems build themselves?

How do we convert systems into proper primitives for input to and output
from a deep multi-processor system?

Suggestions:  convert shapes and images into frequencies using Fourier
transforms.   This is probably as obtuse as a weighted network.

Do we need chatGPT to reason about the systems we might create?

It seems like X3D could lead the way of reasoning about volumes, surfaces,
meshes and vector graphics.  Indeed, I’ve not heard of an updated schema
for SVG.

John

On Sun, Dec 11, 2022 at 6:59 AM John Carlson <yottzumm at gmail.com> wrote:

> Hey, I figured out what to do to build an AGI (general intelligence), I
> think.  I'll let people in on the solution, drop me a note.
>
> Not crazy this time, just mega difficult.  I think people are working a
> couple of levels up right now.
>
> John
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20221211/1ac6b4d4/attachment.html>


More information about the x3d-public mailing list