Will AI have an immediate impact on DevOps? Experts aren’t so sure
For a world where technology is changing in every second, predicting the future is a tough ask. Every promise in the field of technology that creates much excitement falls short of reality — for every right reason off course.
The coming year will bring a lot of growth for DevOps and AI may have some impact on it. Let’s see what experts think about it who are dwelling in the industry for years now.
Chris Carlson, the Vice President of Qualys believes that DevOps will continue to grow and integrating AI with DevOps will be more acceptable for enterprises of all sizes. Hence, the opportunity to improve the different processes of DevOps are endless.
However, seamless integration of security measures into DevOps (aka DevSecOps) are somewhat missing at this point. Vendors who are offering AI integration service merely focus on endpoint & perimeter security with limited toolsets that really can’t evangelize DecSecOps.
Ransomware against endpoints, organized espionage and application level attacks are on the rise. Enterprises that are integrating AI with DevOps still lack the ability to protect themselves from strong threats. We may even see data breaches against cloud environment applications in near future. To overcome this huge challenge enterprises must practice a good culture of DevSecOps.
Christian Beedgen, CTO of Sumo Logic thinks that AI still hasn’t reached to the level to transform DevOps in near future. Strong predictions and promising claims about the significant impact of AI on DevOps are still not taking place in reality.
Yes, there are many examples where AI has made applications smarter, generated code snippets faster or made developers sharper. Yet, when you look at the maturity level of the enterprises it is clear that only around 50% of DevOps companies offer fully digitalized products. It is definite that there is a lot to catch-up in the coming years despite of all the buzz words.
Will AI Replace Programmers?
Microsoft released a Twitter bot back in 2016 named as ‘Tay’ that was designed to mimic teenage American girls and guess what? The twitter bot was shut down just after 16 hours of its release because it started to post offensive tweets. There are many examples of AI issues like Facebook had to withdraw their bots in 2017, a Chinese chatbot ‘Baby Q’ was shut off and so on.
Armando Solar-Lezama a scholar from MIT truly believes that AI has the potential to automate code development but really can’t replace human developers. He suggests to use AI to automate the strenuous parts of code development so that human developers can focus on the creative & complex tasks.
Around the world, startups have initiated their ventures to develop custom software using Artificial Intelligence (AI). Dev9 is such a company based in Seattle. The company use Java and Javascript and strives to eliminate the tedious parts of code development through AI.
Furthermore, new AI models can learn by themselves and can extract rules from the data fed to them. AI is being strengthened by traditional NLP, rule-based reasoning, knowledge representation and human interaction. These processes are being augmented with deep machine learning and as a consequence, new business apps are emerging in a wide variety of domains.
So, What’s Next?
Regardless of what experts say and believe, nearly 30% of software developers are worried to the fact that AI will eventually replace them. But that doesn’t mean you should start looking for an alternative other than DevOps.
It will take a considerable amount of time before AI can really generate practical & production worthy codes. Moreover, AI is not in a position to understand different business propositions of real world, how humans interact with business products or how business value & product features can be set up.
For now, enterprises need to stick to human programmers for their DevOps business who will deliver real value for the software through creativity and prudent advice.
