GenX & GenAI: Beginning of the End of “Boomer Tech”?
The hinderance to AI is past technologies.
And retiring boomers.
Legacy code is becoming a sticky situation as those who support it start to retire or pass away and the next generation of developers are not keeping up their skills to support old and new coding languages at the same time.
The fall back is outsourcing to support legacy development but with US policies focusing on America first — how can we outsource the support of core legacy systems especially if they run government digital ecosystems when the developers live in Asia-Pacific?
Or what if the future of legacy systems is Generative AI and new college graduates?
I taught a summer program at a local university to teach under graduate twenty-somethings how to be old professionals in legacy programming using generative AI — and they were able to ramp up in weeks as opposed to traditional developers who hone their skills over years and years of practice.
I propose if we do not put legacy systems in the hands of college students powered by Artificial Intelligence, it becomes a ticking time bomb for enterprises.
In the meantime, newer and faster technologies are getting blocked and waylaid by so-called ‘Boomer Tech’.
Core systems that feed data downstream that push to Artificial Intelligence such as core banking — still is throwing a monkey wrench into allowing newer technologies to thrive. Also core legacy systems were built with older business processes in mind — so even if you did rebuild or re-architect a legacy system with newer technologies — will the numbers match? Because legacy systems were built often without documentation or comments and how does an enterprise link old business processes (sometimes 20+ years ago) to synchronize to new ways of thinking?
So when we refer to legacy systems what are they?
COBOL
COBOL (Common Business-Oriented Language) is one of the oldest programming languages, born in 1959. It was widely used for business applications, particularly in the financial industry, and is still used in many legacy systems. OBOL almost had a premature death date with the Y2K bug. What happened? In the early days of computing, memory was expensive. To save space, programmers often abbreviated years to two digits (e.g., 99 for 1999). That means when the year 2000 arrived, computers might have interpreted 00 as 1900, leading to incorrect calculations and potential system failures. COBOL has no signs of stopping but most of the support and coding skills exist overseas in Asia-Pacific. Still alive but the developers who support it — getting fewer in number within North America but development thrives overseas especially in Asia-Pacific.
Fortran
Fortran (Formula Translation) is another early programming language, developed in the 1957. It was primarily used for scientific and engineering applications and is still used in some high-performance computing environments such as:
- Aerospace and Defense: Companies like Boeing and Northrop Grumman use Fortran for simulations and analysis.
- Government and Research: Organizations like NASA, national laboratories (e.g., Sandia, Los Alamos), and weather agencies rely on Fortran for research and modeling.
- Energy: Companies in the oil and gas industry use Fortran for reservoir modeling and analysis.
Still alive but in very specific niche markets.
Assembly Language
Assembly language is a low-level programming language that is specific to a particular computer architecture. It is sometimes used to optimize performance or to interact directly with hardware. Assembly allows for low-level control and efficiency and are used for the following instances:
Embedded Systems
- Microcontrollers: These tiny, specialized computers often require code thats highly optimized for size and performance. Assembly Language allows developers to directly manipulate hardware and achieve the tight control needed in devices like: Automotive systems (engine control, braking systems)
- Consumer electronics (remote controls, appliances) Industrial automation (robotics, manufacturing equipment)
- Firmware: The low-level software that initializes and controls hardware components often includes Assembly Language for critical tasks.
Operating Systems
- Kernel Development: The core of an operating system, responsible for managing resources, sometimes uses Assembly Language for tasks like: Hardware initialization Interrupt handling
- Context switching Device Drivers: These programs that allow the OS to communicate with hardware may include Assembly Language for direct hardware access.
A key language to learn for engineering students as newer technologies require maximum hardware performance.
SAS
Statistical Analysis Systems began in 1966 at North Carolina State University. The project began when the university re-hired Anthony Barr to program his analysis of variance and regression software to run on IBM System/360 computers. This was originally intended to analyze agricultural data to improve crop yields. While some sources mention 1971 as the release year, it seems the first official release of SAS was in 1972. However, it was already in use at various universities before this date. The answer to its birthdate might be in how SAS determines dates: SAS datetime values, which store both date and time information, are based on the number of seconds that have elapsed since midnight on January 1, 1960.
SAS code is used for data analysis, statistical modeling, and business intelligence. SAS solutions are implemented across -
- Finance
- Healthcare
- Retail
- Government
- Manufacturing
Could Gen X technologies take over for ‘Boomer Tech’?
Technologies that were created just before and during the dot com boom in the early and late 1990s are still pushing the technology limits of systems today.
But do the following technologies have the staying power such as COBOL and SAS -
- C++ born in 1985 is still alive and kicking. Also making a comeback.
- Python has become the life blood of data science, analytics, and artificial intelligence including generative AI. Python was born in 1989 but is seen as the language that will take over where COBOL left off.
- Java was initially developed by a small team of engineers at Sun Microsystems, led by James Gosling, and was first called Oak. So, while it wasnt officially released until 1995, the birth of Java as a project happened in June 1991. Java is still widely used and considered one of the most popular programming languages today, consistently ranking in the top 5.
- R programming language. R code’s origins can be traced back to August 1993, when it was first conceived by Ross Ihaka and Robert Gentleman at the University of Auckland, New Zealand. R was heavily influenced by the S language, another statistical programming language developed at Bell Labs. Ihaka and Gentleman wanted to create an open-source implementation of S that would be more accessible and widely available. It is has become the primary statistical programming language of choice that is taught globally.
- Julia programming language — not as heavily used as Python and R but built to use both sets of packages. The Julia programming language was publicly released in 2012. It was created by Jeff Bezanson, Stefan Karpinski, Viral B. Shah, and Alan Edelman, a team of researchers at MIT. They set out to create a language that combined the speed of C with the ease of use of Python, aiming to address the shortcomings they saw in existing languages.
Most enterprise companies have been running outdated technologies at the heart of their systems for decades. Now companies are beginning to realize that time is running out to find persons who understand or can migrate these technologies to newer solutions.
I believe using Generative AI to translate ‘Boomer Tech’ into these ‘Gen X’ technologies we can make systems more efficient, faster, and move towards a real-time pace of innovation and creativity. And it shouldn’t take years or cost millions to migrate.
A Quick Win in Modernizing: SAS Alternative
With legacy systems such as SAS that runs on an annual rental model — where you never own the software — enterprises have to find ways to migrate or switch in less than a year or they have to pay for two systems to incur savings. The alternative is to switch to a non open source, black box cloud environment called SAS Viya.
Due to winning a lawsuit, the only alternative to running SAS code in a production environment is a company called Altair who offers SLC — SAS Language Compiler.
This allows for modernizing and generating savings immediately.
Companies can immediately switch over (in less than 6 months) and cuts the software license costs in half. Combine this alternative with Generative AI built in and brand-new university graduates, enterprises can modernize within a year and have trained developers assessing legacy SAS language in order to re-write the code with Gen X alternatives such as — Python, R, and Julia.
SAS is only the beginning. Not only can we modernize SAS code, we can modernize COBOL, JCL, RPG, FORTRAN and many, many other technologies.
With the help of Artificial Intelligence, the next generation of developers will be weaving together old and new to develop flexible and efficient architectures that hopefully soon will allow AI to mutate, migrate on its own based on need. Or AI to suggest new business processes based on new demands or predictive analytic results it gets back.
The future of past technologies is AI.
And hiring new college graduates.
