Episode: Chatting with GPT – How Humans Talk to Machines

Nor Amir S. Abdul Madjid // 2022-08518

1. What language do machines use? And how does it work? (look: binary system)
● Binary number system, in mathematics, positional numeral system employing 2 as
the base and so requiring only two different symbols for its digits, 0 and 1, instead
of the usual 10 different symbols needed in the decimal system. The numbers from
0 to 10 are thus in binary 0, 1, 10, 11, 100, 101, 110, 111, 1000, 1001, and 1010. The
importance of the binary system to information theory and computer technology
derives mainly from the compact and reliable manner in which 0s and 1s can be
represented in electromechanical devices with two states—such as “on-off,”
“open-closed,” or “go–no go.” (See numerals and numeral systems: The binary
system.
● Machine language, the numeric codes for the operations that a particular computer
can execute directly. The codes are strings of 0s and 1s, or binary digits (“bits”),
which are frequently converted both from and to hexadecimal (base 16) for human
viewing and modification. Machine language instructions typically use some bits to
represent operations, such as addition, and some to represent operands, or
perhaps the location of the next instruction. Machine language is difficult to read
and write, since it does not resemble conventional mathematical notation or human
language, and its codes vary from computer to computer. [1]
● A binary code represents text, computer processor instructions, or any other data
using a two-symbol system. The two-symbol system used is often "0" and "1" from
the binary number system. The binary code assigns a pattern of binary digits, also
known as bits, to each character, instruction, etc. For example, a binary string of
eight bits (which is also called a byte) can represent any of 256 possible values
and can, therefore, represent a wide variety of different items. [2]

2. If machines only know 1s and 0s, how can we humans talk to them? (look: assembly
and programming languages)
● Assembly language, type of low-level computer programming language consisting
mostly of symbolic equivalents of a particular computer’s machine language.
Computers produced by different manufacturers have different machine languages
and require different assemblers and assembly languages. Some assembly
languages can be used to convert the code that programmers write (source code)
into machine language (readable by the computer) and have functions to facilitate
programming (e.g., by combining a sequence of several instructions into one
entity). Programming in assembly languages requires extensive knowledge of
computer architecture. [3]
● Assembly language is one level above machine language. It uses short mnemonic
codes for instructions and allows the programmer to introduce names for blocks of
memory that hold data. One might thus write “add pay, total” instead of
 “0110101100101000” for an instruction that adds two numbers. Assembly language
is designed to be easily translated into machine language. Although blocks of data
may be referred to by name instead of by their machine addresses, assembly
language does not provide more sophisticated means of organizing complex
information. Like machine language, assembly language requires detailed
knowledge of a particular internal computer architecture. It is useful when such
details are important, as in programming a computer to interact with input/output
devices (printers, scanners, storage devices, and so forth). [2]
3. How did we go from binary communication with machines to the human-like computer
conversations we have today? (look: natural language processing)
● Natural language processing (NLP) is an interdisciplinary subfield of computer
science and linguistics. It is primarily concerned with giving computers the ability
to support and manipulate speech. It involves processing natural language
datasets, such as text corpora or speech corpora, using either rule-based or
probabilistic (i.e. statistical and, most recently, neural network-based) machine
learning approaches. The goal is a computer capable of "understanding" the
contents of documents, including the contextual nuances of the language within
them. The technology can then accurately extract information and insights
contained in the documents as well as categorize and organize the documents
themselves. Challenges in natural language processing frequently involve speech
recognition, natural-language understanding, and natural-language generation. [4]
● Natural language processing (NLP) refers to the branch of computer science—and
more specifically, the branch of artificial intelligence or AI—concerned with giving
computers the ability to understand text and spoken words in much the same way
human beings can.

NLP combines computational linguistics—rule-based modeling of human

NLP drives computer programs that translate text from one language to another,
respond to spoken commands, and summarize large volumes of text rapidly—even
in real time. There’s a good chance you’ve interacted with NLP in the form of
voice-operated GPS systems, digital assistants, speech-to-text dictation software,
customer service chatbots, and other consumer conveniences. But NLP also plays
a growing role in enterprise solutions that help streamline business operations,
increase employee productivity, and simplify mission-critical business processes.
[5]
4. Is machine language always binary at the lowest level? (look: ternary system and
quantum bits)
 ● A ternary computer (also called a trinary computer) is a computer that uses ternary
logic (three possible values) instead of the more common binary logic (two
possible values) in its calculations.
● Ternary is the new binary. Classical computing favors 0's and 1's mainly because of
practical engineering. Using three states - ternary - is mathematically proven to be
the optimal discrete signal encoding on the lowest hardware level. Three states will
allow computers to think beyond black and white and introduce new fundamental
logical constructs to reason with. Imagine having an extra letter in the alphabet
and the new words that can be made. Compared to binary, data will be more
compressed, faster processed, have less latency in communication and use less
energy. [6]
● If ternary computing is so much more efficient, why aren’t we all using them? One
answer is that things just didn’t happen that way. We have come so far in binary
digital computing that it would be pretty hard to turn back. Just as the robot
Bender has no idea how to count beyond zero and one, today’s computers operate
on a logic system that’s different from what any potential ternary computer would
use. [7]
● Quantum computers do calculations using bits, too. After all, we want them to plug
into our existing data and computers. But quantum bits, or qubits, have unique and
powerful properties that allow a group of them to do much more than an equivalent
number of conventional bits. Qubits can be built in various ways, but they all
represent digital 0s and 1s using the quantum properties of something that can be
controlled electronically. Popular examples—at least among a very select slice of
humanity—include superconducting circuits, or individual atoms levitated inside
electromagnetic fields. The magic power of quantum computing is that this
arrangement lets qubits do more than just flip between 0 and 1. Treat them right
and they can flip into a mysterious extra mode called a superposition.

You may have heard that a qubit in superposition is both 0 and 1 at the same time.
That’s not quite true and also not quite false. The qubit in superposition has some
probability of being 1 or 0, but it represents neither state, just like our quarter
flipping into the air is neither heads nor tails, but some probability of both. In the
simplified and, dare we say, perfect world of this explainer, the important thing to
know is that the math of a superposition describes the probability of discovering
either a 0 or 1 when a qubit is read out. The operation of reading a qubit’s value
crashes it out of a mix of probabilities into a single clear-cut state, analogous to
the quarter landing on the table with one side definitively up. A quantum computer
can use a collection of qubits in superpositions to play with different possible
paths through a calculation. If done correctly, the pointers to incorrect paths
cancel out, leaving the correct answer when the qubits are read out as 0s and 1s.
[8]
5. What do you think is the future of computer language?
 ● Quantum Computing

[1] https://www.britannica.com/technology/machine-language
[2] https://en.wikipedia.org/wiki/Binary_code
[3] https://www.britannica.com/technology/assembly-language
[4] https://en.m.wikipedia.org/wiki/Natural_language_processing
[5] https://www.ibm.com/topics/natural-language-processing
[6] https://ternaryresearch.com/
[7] https://www.techopedia.com/why-not-ternary-computers/2/32427
[8] https://www.wired.com/story/wired-guide-to-quantum-computing/