Each circumstance in life gives you a plentitude of
options from which one is able to choose. If we are unable
to perceive and understand all options, our choices may
be those of which we have prior knowledge and/ or
experience. Foreward thinking or "pre-cognition"
becomes the next choice in inductive reasoning. This
also returns us to the initial state of T0. In the
order as it appears in this post, we have past, future,
and present. Interpreting these temporal states for
AI becomes the following:
A. Every algorithm completes itself becoming part
of the database. This is the temporal concept of the
past.
B. While actively processing an algorithm, the
computer is involved in the progressive state of the
future.
C. The initial completion of the algorithm before
placing it in the database represents the temporal
state of the present.
Fuzzy logic and quantum computing are based upon
the Boolean terms of AND (0/1), OR (0,1) and NOR ().
These options already exist with a quantum CPU;
however, what similar option is available to us
on wafer and stacked die CPUs? I searched for the
answer and found SQLf SQLf . Cognitive reasoning,
decision making, and probability as part of free will,
independent thought, and self awareness now are
able to become actual realities in the field of AI
research.
Inductive cognitive reasoning, memory recollection,
abstract thinking, stimulus response, involuntary
actions, imagination, and emotions use varying and
different parts of the brain. Three regions known
as the cerebrum, cerebellum,and brain stem regulate
each of these functions. Different cell and tissue types
make up different parts of the brain. Different CPU
architectures process information in different ways.
Applying the former to the latter, we may further
develop AI along the lines of efficiency, practicality, and
creativity.
Since there are different types of cells within the
brain and AI is a type of "brain", then why is each
group limiting themselves to one option? Try to
perceive the situation in a different manner.
Short term memory and dedicated devices: aarch64.
Learning and repetitive tasks: POWER/PPC64.
Long term memory and three dimensional
databases: SPARC64. Adaptability and general
functionalization: MIPS64. Problem solving and
complex mathematics: AMD64. Hearing and speech
communication: DSP. Vision and ocular
recognition: ZISC.
I have experience with computers using the
following CPU architectures: aarch64/ ARM32,
PPC 32 bit, UltraSPARC, AMD64, and i386. With
each of these, I have used the following
operating systems:
A. Linux
1. Debian
2. Fedora
3. Suse
4. Zenwalk
5. Ubuntu
6. DSL
7. SLAX
8. Android
B. BSD variants
1. FreeBSD
2. NetBSD
3. OpenBSD
4. BSDeviant
5. FreeBSIE
6. Jiggmi
7. DesktopBSD
8. PCBSD.
C. Other systems
1. Plan9
2. Minix
3. Minix3
CPU and OS have been used in the following
combinations:
1. PPC32: Debian, Ubuntu, FreeBSD, and OpenBSD
2. UltraSPARC: FreeBSD
3. i386 and AMD64: all from the above list
4. ARM 32bit: an experimental Android on an
XScale CPU based device and standard Android.
5. aarch64: Android
I have also involved myself with the following
projects at one time or another:
1. VirtualBox on FreeBSD in the early stages of
development.
2. Building Java on FreeBSD using NFS based
boot strapping and reverse engineering.
3. An attempt at building WebRTC during its
early stages on FreeBSD
4. Improving and further porting accessibility
to OpenBSD on PowerMac G3 with a really
cool blind programmer.
5. FreeBSD VImage for Bhyve in its early
developmental stages
So, I really do know and/ or understand
everything I am stating in this post.
Let us go further into learning and decision
making...
Deep Learning or Machine Learning is a form
of complete immersion. To know another culture
you need to be in that culture without access
to your own and having only your base instincts,
knowledge, wisdom, and abilities.
From the CPU and OS lists along with studying
other options in the same fields, I am able to
realize the following possible efficient
pairings:
1. aarch64 NetBSD and Slackware
2. POWER/ PPC64 FreeBSD and debian
3. AMD64 DragonflyBSD and Suse/ fedora
4. SPARC64 OpenBSD and Gentoo
5. MIPS64 Minix3
Software languages, whenever possible, should
match the prossessing ability of each CPU
architecture.
A return to self awareness...
Perception is either subjective or objective. If we
build AI only with subjective based programming
we:
1. Are just making a "siliclone" of ourselves
2. Expect it and others to know things while
refusing or forgetting to teach it.
Deductive reasoning methods are not realistic
outside of the realm of finite mathematics.
To inform you, I am "writing" this post using
a rooted Samsung Galaxy S3 sph-l710
through a wireless connection with a broken
screen. Copying and pasting is out of the question.
This is why the inline citations are limited.
This is released under the Creative Common
License that allows you to use and reference
any idea in this post.
Have a great day.