Audio version of the article
Predicting the next developments in DevOps has become a popular pastime for many developers. We’ve seen our industry change rapidly over the past decade, and the role of the programmer has changed dramatically over the same period, and one can say due to artificial intelligence.
In fact, and especially among developers from (say) a certain era, it seems that the role of the “traditional” developer no longer exists. Rather than the traditional software development life cycle – in which software is planned, built, and then released – many of us now work in multi-disciplinary teams in which development and operations exist side-by-side, and are increasingly indistinguishable.
This model came to be known as DevOps, but only recently. And it only took DevOps a few years (or months?) to start making changes and adjustments. Now we’re floating in a sea of acronyms (including the latest DevSecOps) denoting a slightly different way of working, each with a slightly different vision of the future.
One of the newest is AIOps. In this vision, AI tools are slowly replacing the developer role, just as DevOps did before, and will eventually replace DevOps entirely.
Judging whether this prediction is true is difficult, but in this article we’re going to try anyway. We’ll see what AI promises for the development process, we’ll assess whether it can really replace human developers, and then we’ll see what DevOps is likely to look like in a few decades.
The Promise of Automation
First, and to understand why DevOps teams adopted AI tools so quickly, it helps to understand what they promise, what they can achieve. This promise can essentially be divided into two parts:
- On the one hand, the use of AI tools is “only” one way for revised developers to keep pace with increasingly complex systems.
- AI tools, on the other hand, offer a variety of automated code development and implementation techniques that have fundamentally changed the way software is created.
Regarding the former, the AI tools that make our lives easier, we just need to look at the prevalence of cloud infrastructure models over the past decade to see why AI was necessary. Infrastructure, and its management, has become almost impossible without artificial intelligence.
In fact, hybrid and multi-cloud infrastructures, microservice architectures such as containers and hyperscale applications have created a corporate IT environment that is more complicated than ever to control. Then there is the second aspect of introducing AI tools: the fact that some AI tools like GitHub’s AI Coding Assistant or Microsoft’s DeepDev not only make our lives easier, but also open up new possibilities for development and distribution of code open up.
The speed with which artificial intelligence systems can record the requirements of individual users, for example, makes it possible to adapt and set the default behavior when surfing the Internet for each user individually, thereby reducing their susceptibility to certain types of malware while using our software.
DevOps, AIOps, and NoOps
Because of the usefulness of artificial intelligence tools, they have been widely and quickly adopted by all DevOps teams. AI interfaces have almost become a necessity as you evolve and scale your DevOps program.
The most obvious and tangible result of this change has been the data and systems developers spend their time observing. For example, it used to be an important part of the operations team’s tasks to set up and maintain a dashboard that all employees could access and that contained all relevant data in one piece of software.
Today this core task is largely obsolete. As software has become more complex, the idea of a single dashboard that contains all of the relevant information about a particular software seems absurd of artificial intelligence tools that “automatically” monitor the software they are working on and only present data when it is clear that something has gone wrong.
- This is a massive change in the way we work as developers and operations staff, so it’s no wonder it got its own acronym: AIOps.
- Some have gone even further, claiming that this reliance on artificial intelligence tools now means we are entering the era of NoOps.
- However, opinions are clearly divided as to which feature has been removed, whether NoOps means “no developers” or “no operation”.
The AI revolution
All these considerations on the philosophical background of software development may sound a bit abstract: Speed (if not quality) with which the software is delivered.
GitLab’s latest survey of more than four thousand developers provides some tough numbers. This research found that some companies are releasing new code up to ten times faster than before. Significantly, 75% use AI and ML to test and review the code preview, up from just over 40% a year ago.
This is great news for developers, or at least those who want to create a large amount of code quickly. Unfortunately, that doesn’t always mean great code. For example, ransomware attack trends show that poorly tested code can quickly become a problem. A notable vulnerability for many organizations, and the advent of AI-powered test systems has done little to alleviate it.
The direction of travel is surprisingly clear. In a few years, it seems that the vast majority of DevOps teams will be turning to AI tools, with the software being rolled out much faster than before. That brings us back to the question we started with: With AI tools doing so much DevOps work now, do we still need human DevOps staff?
Well yes and no. Find the reason.
At the most superficial level, this question can be answered with a very simple test. Have developers noticed a decrease in their workloads with the introduction of AI tools across the development industry and you will be greeted with a laugh. The answer is no.
Because DevOps teams spend less time managing their software on a day-to-day basis, the time they used to spend on it is now being spent on potentially more valuable tasks: strategic planning, meta-analysis and ensuring your development goals are development online.
Indeed, the AI ”revolution” that many thought would make DevOps obsolete seems to have made teams bigger and busier than ever. This is because it had the side effect of making development cycles much faster.
This is well understood by industry managers. In a recent interview with ZDNet, Matthew Tiani, Executive Vice President of iTech AG, stated that DevOps is now being improved among other things through improved source code management of the technical tool set, CI / CD, orchestration.
He added that the successful implementation of DevOps uses a compliant development methodology such as Agile and Scrum, and a commitment by the organization to encourage and encourage collaboration between development and operations personnel.
Learning to Live Together
In practice, these factors and trends mean that DevOps teams are increasingly focused on business goals rather than technical challenges. This is certainly a change, but it may not be a negative in terms of the quality of the software. In fact, AI tools have given teams the opportunity to focus their human resources where they work best: on creative, holistic and strategic tasks, for example that Vue JS is now the fastest development framework in the world with more than 240,000 System built websites are now live all over the world it would be ridiculous.
Thanks to artificial intelligence tools, DevOps teams now have much more freedom in their decisions and are confident that their tools are advanced enough to overcome technical challenges.
This means that AI tools will most likely not replace DevOps, or at least in the foreseeable future. As we’ve seen with previous trends in adopting DevOps, development teams still need strategic leadership no matter how advanced their technical tools are. And while AI is very powerful in many areas today, it still lacks the ability to respond to user demand in a really creative way.