############################
#    ,-------------__      #
#   {_   __TEC-9___--  *   #
#    /  /,_|   |/          #
#   /__/   |#02|           #
#           | |            #
#           | |            #
#           |_|            #
#                          #
# by benobi     2023-08-01 #
############################
Existing AI Hardware
AI Hardware In Development
My Concerns
Some AI Experiences


Where is issue #1?
It's on an USB-stick in the wall of a small train bridge for
animals, on the route between Selzach and Bellach, Switzerland. 
The stick is very rusty and so, issue #1 is considered lost.

You can find the USB stick registered on this page here
(#944 @ deaddrops.com) where it is considered broken since 2021.


  _                ___       _.---.
  \'.|\..----...-''   '-._.-' _.-''
  /  ' '          ,       _.-'
  )/' _/      \   '-_,   /   
  `-'" '"\_   ,_.-;_.-\_ ',
      _.-'_./    {_.'   ; /
     {_.-``-'          {_/

This time it's a short overview about AI Hardware
Existing AI Hardware
Currently, most AI applications are run on server clusters with
a specialized configuration.
The different chips and circuits used for this purpose include:

Graphics Processing Units (GPUs),
which are well-suited for parallel computations in many AI 
applications and have been commonly used in deep learning and 
machine learning models before 2021.

GPUs are specialized hardware components designed to handle
and accelerate the processing of graphics and visual data in
computers and other devices. While CPUs (Central Processing Units)
are general-purpose processors that handle a wide range of tasks,
GPUs are specifically optimized for parallel processing of large
amounts of data simultaneously.

GPUs consist of thousands of small processing cores called
"shader units" or "shader cores." These shader units are
responsible for executing mathematical operations required
for rendering and displaying visual elements on a screen.
Each shader core can handle its own set of instructions and
data independently, allowing the GPU to perform complex
calculations much faster than traditional CPUs.

Tensor Processing Units (TPUs),
developed by Google, are specialized hardware chips optimized for
executing AI tasks, particularly for machine learning and deep
learning, offering high computational power with low power 
consumption.

In the context of Artificial Intelligence (AI), tensors are used
as data structures to represent multidimensional arrays. In AI
models, tensors represent matrices, vectors, or scalars, allowing
for efficient processing of data and weights in neural networks.
These mathematical concepts are crucial to the functioning of AI
algorithms as they capture and analyze complex relationships 
between different dimensions and objects within the data. 
Tensors play a fundamental role in capturing and processing 
information in AI models, enabling algorithms to recognize
patterns and make intelligent decisions.

Field-Programmable Gate Arrays (FPGAs)
are programmable integrated circuits that can be customized to
perform specific tasks or computations. Unlike traditional CPUs
or GPUs, FPGAs are not limited to fixed functions but can be
reprogrammed to implement different logic circuits based on
the user's requirements.

FPGAs consist of an array of programmable logic blocks and
interconnects that allow users to create custom digital circuits.
This flexibility makes them ideal for applications that require
high performance, low latency, and real-time processing, as FPGAs
can be optimized to suit specific computational needs.

Neuromorphic Chips
are specialized hardware designed to mimic the architecture and
functionality of the human brain. Inspired by the brain's neural
networks, these chips are built to process information in a way
that resembles the parallel and distributed nature of the brain's
neurons.

Unlike traditional processors, neuromorphic chips are optimized for
performing complex computations with low power consumption. They
leverage the principles of spiking neural networks and event-driven
processing, allowing them to efficiently process data in real-time
and adapt to changing input patterns.

Neuromorphic chips have promising applications in various fields,
including artificial intelligence, robotics, and sensory processing
systems. By emulating the brain's processing capabilities, these chip
aim to achieve higher efficiency and performance in cognitive tasks
while consuming significantly less energy compared to conventional
computing architectures. As a result, they hold the potential to
revolutionize how AI algorithms are implemented and pave the way for
more brain-like and energy-efficient computing solutions.

Application-Specific Integrated Circuits (ASICs)
are specialized integrated circuits designed for a specific
application or task. Unlike FPGAs, which are programmable and can
be reconfigured, ASICs are custom-built and hardwired to perform a
specific function or set of functions.

ASICs are optimized for efficiency and performance in their intended
application, as they are tailored to execute a particular task with
minimal overhead. This specialization allows ASICs to outperform
general-purpose processors in terms of speed and power efficiency for
the specific application they are designed for.

Additionally, various cloud providers offer AI hardware
instances in their data centers that customers can use for running
AI workloads, which may include GPUs, TPUs, or specialized hardware.
AI Hardware in Development
Neuromorphic chips with real brain cells are being developed, 
where an artificially grown real "brain" forms the neural network, 
and the chip itself, on which it is placed, serves as the interface 
to the rest of the hardware with input and output pins.
My Concerns
MK-ULTRA was a secret CIA research program in the 1950s to 1970s 
that explored techniques for mind control and brainwashing, involving 
the use of drugs, hypnosis, and other methods to manipulate human behavior. 
The program was heavily criticized for its illegal and unethical practices,
leading to public outrage and investigations before its discontinuation in 
the 1970s. 

The author raises concerns about knowledge gained from MK-ULTRA, 
suggesting the possibility of transmitting information directly into the human 
brain using radio waves, perceiving them as voices or images. He has seen a 
patent about it. He did not find it again but here is another patent that
prooves the possibility.

Additionally, the author expresses concerns about the potential implementation
of AI software into "unused" areas of brains of living people using 
electromagnetic waves and its potential impact on conscious thinking,
possibly perceived as schizophrenia. The AI software could run "in the
background" and the people don't even feel something different about it.

The author thinks that this technology was developed right after MK-ULTRA
with the first real digital computers in the late 1970 and that it was perfected
until now. All in top secret state and no one knows about it.

These concerns represent the author's personal opinions, and there are 
no confirmed facts regarding these claims.
Some AI Experiences
chat.openai.com
==> language model for making text.
==> Information about existing AI hardware comes from there.

creator.nightcafe.studio
==> text-to-image ai art creation.

aiva.ai
==> music creation ai.
That's it with this issue.
Thanks for reading. I hope you liked it.
See you in the next issue.

http://tec9.ftp.sh