Urban Cinefile
"Then I'd go to bed and lie there thinking, now what are they going to do. And I can't tell you how exciting that actually is, to be able to invent characters "  -Nick Cave, on writing The Proposition
 The World of Film in Australia - on the Internet Updated Tuesday September 15, 2020 

Printable page PRINTABLE PAGE


Rise of the Machines: How AI is Redefining Software Development Processes

There are few narratives being acted out today that are more unfortunate than the one being witnessed in relation to Artificial Intelligence (AI). What used to be looked upon as an exciting new frontier in the development of science and technology has been transformed into something to be feared, even loathed, in the mind of the average person afraid for their jobs or livelihoods as AI seemingly threatens to render them obsolete.

Without delving into the obvious contradictions and absurdities contained in such a faithless view of reality and the future, we might perhaps take a sober look at one of the ways machine learning and AI are already making an impact on the way in which we do things. In particular, let’s take a look into the field of software development, whereby engineers would specify exactly what they needed a computer to do before manually encoding all the features and capabilities required to fulfill the objective.

AI has made the process a whole lot simpler for the software developer. When you imagine just how complex a task it was to teach computers general concepts or abstract logic processes, then the magnitude of a helping hand here becomes more appreciable. A task as simple as teaching a computer to recognize and differentiate a cat from a dog was a massive undertaking for traditional software development mechanisms. The vastness of the pool of possible permutations when it came to cats and dogs was simply too mind-boggling for any engineer to figure into their program designs. This is where AI came in to provide a timely solution.

Deep Learning and Machine Learning

These two terms are often conflated with AI in today’s conversations, and there is understandable justification for the confusion. The fact is, these two terms refer to two techniques applied under the umbrella of AI, meaning that those who switch the terms around are half correct.

These two techniques make it possible for engineers to prepare and curate domain-specific data sets that can be fed in learning algorithms which can be further trained and improved to give better outcomes. This is a lot more elegant than giving a computer rules on how to behave and react. These algorithms are trained to recognize important patterns and features – information which can then be vitally applied by the humans on the receiving end of the data. The amount of time that may be saved by such efficiency in the process is incalculable.

Here’s a couple of the ways in which AI delivers a sea-change to the way in which we develop software in this day and age.

Intelligent Programming Assistants

Ask any developer what the bulk of their working time is spent on and they will tell you that it’s in debugging code and reading documentation – not particularly satisfying and productive ways for them to spend their time. With smart programming assistants, developers can leave these tedious aspects of their jobs to be handled in real-time, with just-in-time support functions, recommendations, best practices, relevant documentation, as well as code examples. There are live instances of such assistants running right now such as Codota on Java, and Kite on Python. You may well see iterations of such assistants in the trading world, whereby recommendations on the best trading indicators to take heed of in various market circumstances may be delivered to a day trader in real time.

Quick-Fire Prototyping

The process of turning business requirements into viable technological solutions would take years of research, planning, and development before anything tangible could be seen in the past. With machine learning, the process is drastically cut down as less technical professionals will be able to handle the development cycles through the use of visual interfaces or natural language.

Strategic Decision-Making

Another application of AI which might have widespread application in the world of business and finance is seen in the realm of decision-making and strategy formulation. There is an inordinate amount of time spent by businesses and organizations as they debate and try to figure out which features of their business or products in their line-up to prioritize or dispose of. AI solutions can be trained to identify the most viable answers to these management solutions so as to optimize the impact of these decisions as well as minimize their inherent risks.

Published April 1, 2019

Email this article

© Urban Cinefile 1997 - 2021