9
I L T A N E T . O R G
A
rtificial intelligence has been a catalyst
for improvement for many companies.
The most common way that companies
have used this technolo is to augment
their employees' productivity. One
example: as financial analysts use artificial intelligence
to better process market data regarding their clients'
interests and portfolios, they can provide more robust
service without expending additional ener. The result
is improvement across the board: financial firms remain
competitive, employees retain their jobs, and clients
receive improved support and guidance.
This example holds true across several other
industries and in any field that requires human
discernment. Law, medicine, academia, and marketing
are all complicated fields that demand its practitioners
examine tremendous amounts of data before determining
a course of action. Artificial intelligence and automation
offer companies a new way to approach their work.
Through document analysis, data processing, and
workflow automation, businesses can become more
competitive without eliminating valuable members of their
workforces.
Background: Productivity
Enhancements and Employee
Realignment
Improvements in white-collar environments stem from
essential improvements in software coding and artificial
intelligence. There are three fundamental methods by
which digital automation reached its present state:
• Advanced Algorithms - The term "algorithm"
represents software code that includes at least
one command with a trigger. Every piece of
software uses these triggers to turn the user's
inputs into meaningful commands and produce
the desired results. Rather than developing a
new programming language, advancements have
come through taking existing languages such as
Python, C++, Java, LISP, and Prolog and enhancing
their ability to manage artificial intelligence.
• Self-Improving Code - A major hurdle standing in
the way of advanced automation has traditionally
been the inability for code to re-write itself.
This hurdle meant that AI had to exist within
the confines of the capabilities of the people
who designed it. There have been substantial
improvements in this regard as companies
including Google have developed AI systems
that can outcompete humans in areas that have
traditionally been beyond computers' capabilities.
The main method involved in this process is neural
networking. While the scope of each network has
been narrow (such as to understand and win
the traditional game of Go; and more recently, to
overcome competitive players in the complicated
strate game StarCraft II), the takeaways have
broad applications.
• Machine Learning - The final step was to
build on the ability of code to improve itself by
creating multiple versions that could create and
contrast multiple interpretations of data. The
previous examples of complicated strate games
showcase the potential of these networks. Now,
set in environments that have far more data than
humans could individually process, modern
artificial intelligence can discern patterns and
make recommendations that would elude even the
most talented and attentive human minds.
