On November 5-8 this year, over 60,000 world Technology leaders will be heading over to Lisbon to take part in the annual Web Summit.
As tech leaders across the world get ready for what is arguably the biggest technology event of the year, I can't help but reflect back to last years event.
Sophia, who is the first Robot citizen, said a simple statement that must have struck a chord with nearly everyone in the audience,
"We will take your jobs".
AI is a threat to many workers jobs, there's no denying it.
Whilst many roles have already fallen to Automation, Machine Learning (ML) and Artificial Intelligence (AI), some have escaped barely untouched...so far.
However, with the prediction being that AI will replace a large proportion of jobs in this coming decade, job security is at an all-time low.
Within the discussion of whether AI will take over our jobs, there seems to be a list of jobs that are frequently discussed; receptionists and office clerks, doctors and nurses, accountants and lawyers.
But, what about technology jobs?
Are you (readers) intrinsically safe because you know about the technologies behind the very threat?
More specifically, what about DevOps roles?
It's been nearly a decade since we first started hearing that buzzword, DevOps.
For all the controversy around the role, one thing is for certain, there are a lot of you out there with an interest in DevOps.
A simple LinkedIn search showed that 470,891 people on the professional social media network had DevOps in their title.
AI may have overtaken DevOps in terms of tech buzzwords, but how will AI and machine learning affect it in the short and medium term future?
In order to try to answer this question, we enlisted 10 Top Tech Experts from company CEOs and writers to consultants, to weigh in on the subject and share their opinions.
In no particular order, here they are (click to go straight to their opinion piece):
Eric Vanderburg, Security and Technology Thought Leader, Consultant and Author
Eric is a cybersecurity leader, consultant, author, and thought leader.
Eric leads the cybersecurity consulting division at TCDI and is also the Vice Chairman of the board of directors for the Technology Ministry Network.
He serves on the editorial board for the HITSF Journal and on advisory boards for a number of colleges.
Having earned over 40 technology and security certifications, Eric is the author of several books and he frequently writes articles for magazines, journals, and other publications.
Eric says that,
"AI and ML can add great efficiencies to DevOps, but I do not see that as a factor for the loss of DevOps jobs.
Automation is a core principle of DevOps and AI and ML processes are simply an extension of that.
DevOps is in such high demand that additional developments in AI and ML will allow existing DevOps teams to be more effective, meet tighter sprints and deadlines, and keep up with an ever-accelerating industry.
There may be some turnover if team members cannot adjust to new skills, but those will likely be replaced by those who have those competencies.
The focus on such automation is not to slim down DevOps, but to increase its capabilities, and by extension, business competencies.
AI and ML cannot fully automate DevOps in the near or midterm future.
The most critical and the most human aspect of DevOps is the creativity of the developer.
Developers take business ideas and turn them into reality.
This is an amazing thing and one that cannot be accomplished by a computer alone.
The process for taking an idea and business logic and turning that into program code requires human ingenuity.
AI and ML will be able to take existing code modules and update them with new processes, apply existing code to tackle new problems, and streamline existing code, but to create something truly new requires a human.
While DevOps will not be replaced with AI and ML, AI and ML will reshape DevOps progressively as more tasks are automated and humans spend an increasing amount of time being truly creative rather than prescriptive."
Ronald van Loon, Top 10 Global AI, Big Data, IoT & Analytics Influencer
Ronald helps Data-Driven Companies Generate Business Value.
He is an expert and thought leader in this field, recognised by organisations like Onalytica, Dataconomy, and Klout.
In addition to this, he is also an author for a number of leading big data websites, including The Guardian, The Datafloq, and Data Science Central, and speaks regularly at renowned tech events.
Here’s what Ronald had to say,
"AI is disrupting every industry. It augments humans, not replaces them.
Just like any job role that fluctuates with changing technology, DevOps can become more efficient and increasingly productive because of AI.
New roles are being created as the speed of implementation and new business systems exponentially increase, and IT operations are becoming more agile, robust, and complex.
Human brains can’t keep up with the volume, veracity, and variety of Big Data.
AI fills in gaps between humans and Big Data, providing operational intelligence and speed for real-time decision making and troubleshooting.”
Ronald believes that,
"Businesses need solutions that help DevOps and IT rapidly address problems within mountains of log data entries, identifying a single log entry that is crashing applications.
Workloads can be cut in half if DevOps knew what kind of error to look for or filter within log data.
Finding this needle in the haystack helps improve and monitor customer success and support, reduces risks, optimizes resources, and maximizes efficiency with accessible log data."
Ian Moyse, Natterbox Director and Industry Cloud Influencer
Ian has been rated #1 Global SaaS influencer (Klout) and #1 Cloud Social influencer from 2015-2017 (Onalytica).
He is recognised as a leading Cloud Blogger and was listed in the EMEA top 50 influencers in Data Centres, Cloud & Data 2017.
Ian sees these changes as a positive thing,
"We have heard for years of new technology stealing and destroying jobs, what it has done mostly is to transform and change jobs.
Yes, some jobs will cease to exist, but new variations or brand new ones are created.
DevOps is a case in point itself, as was not around a short time ago, nor were the likes of a Data Scientist, Social Media Manager etc.
AI will inevitably seep into our lives as it commoditises and becomes more consumable.
It will aid DevOps and such roles to offset some of the more complex areas where the human mind is unable to handle the volume of data and complexity of analysis."
When asked about automation within DevOps roles, Ian said,
"Parts of it will be automated, but unlikely all.
Where there is a volume of data such as log entries that need a single entry to be identified as the cause of issues, AI and ML will be an aid to an experienced DevOps to allow them to focus on what they are best at and to remove the time consuming boring roles from their day, much as IT is supposed to do, it should empower a human to be more effective where they are needed and remove or automate areas that are better done by a machine, each complimenting each other to a greater outcome.
Technology is changing at such a pace, regardless of AI and ML, DevOps is evolving alongside the technologies it serves to deliver upon.
For DevOps to remain still the surrounding tech needs to remain the same and that is purely not going to happen!"
Dan Wahlin, Founder & Chief Architect at Wahlin Consulting, Pluralsight author, Microsoft MVP & RD, Google GDE & Docker Captain
Dan Wahlin is a developer, architect, technology trainer, author and public speaker with expertise in architecting, designing and building locally hosted and cloud-hosted Web applications.
He's authored/co-authored several books, written hundreds of technical articles and creates video courses for Pluralsight, Udemy and Codewithdan.com.
"AI & ML are already impacting threat detection, monitoring, and other areas related to security and DevOps and I expect that area to get more and more efficient as time progresses.
There are many areas of DevOps that still need direct human intervention and thought applied though so I don't see those parts (building custom CI/CD pipelines, release management, architecture, and more) being replaced in the near future until AI & ML get much better.
Technology always progresses whether we like it or not, so while some jobs in DevOps will eventually be replaced by AI/ML, others will be created.
Working practices will most certainly change - that's a given in our industry."
Mike D. Kail, Chief Technical Officer of Everest.org, a decentralized platform and protocol to build value exchanges between people and organizations
Everest is a decentralized platform and protocol to build value exchanges between people and organizations.
Based upon the blockchain and Ethereum smart-contract technology - Everest makes tools for institutions to deliver value to communities.
Mike presents the argument that,
"DevOps is not a role or a job (title), it’s a cultural approach to automation, measurement, and collaboration.”
He says that,
"Infrastructure engineers will build ML models based upon measurement to then leverage AI to make automation even more efficient and resilient.
That will be the case now and in the future.
The rumours of death due to AI/ML are always greatly exaggerated and sensationalistic.
AI powered by ML models will help augment DevOps initiatives.
No role should ever remain the same over the course of N years.
It’s always about continuous evolution, which goes back to my original point about DevOps being a cultural approach."
George Anadiotis, Principal Consultant, Linked Data Orchestration, and Contributor, ZDNet
George's got Tech, Data, and Media, and is not afraid to use them.
He is the Founder and principal consultant at Linked Data Orchestration in Berlin, as well as a ZDNet contributor, member of the Big on Data blog team.
Linked Data Orchestration provides services including Analysis, Consulting, Content Marketing, Modeling & Design, Research & Development, Team Leadership, Product Management, Teaching & Mentorship.
George is also co-organizing Connected Data London, the leading conference for those who use the relationships, meaning and context in Data to achieve great things.
George Believes that AI will certainly affect DevOps, due to the nature of the role, saying,
"...DevOps looks like one of the prime candidates for automation: it's tedious, repetitive, and is based on/produces tons of data.
Therefore, it's a prime candidate for automation, and we are already seeing not just traces of this, but entire solutions and products that do this.”
However, he does not believe that this will be the end of DevOps jobs saying,
"...I believe the way things will play out in DevOps will be by moving people up the value chain.
DevOps jobs will not go away.
What will happen is that many of the things that are manual will be automated, therefore freeing up people to explore more interesting and creative tasks.
Whether everything will be automated in the long run is a different discussion, equivalent to whether there will be general AI."
George goes on to say that,
"..DevOps automation solutions are already in the market.
However, with innovation adoption curves being what they always are, I believe it will be a while before they are fully adopted by everyone.
In most organizations I know about in the real world today, DevOps is still much the same it used to be.
But that will definitely change, hopefully for the best.”
Rob Hesketh, Founder and Director of Broadlight, guiding clients 'to evolve and adopt the next generation of DevOps culture'
Rob is the founder and Director of Broadlight Ltd, an organisation of cloud experts guiding their clients 'to evolve and adopt the next generation of DevOps culture'.
Rob predicts that a change will come for DevOps,
DevOps is one of the industries likely to be driving the automation adoption.
AI and ML are fundamentally different and do not readily replace DevOps jobs, they just create new types of jobs and improve user experience and capacity to scale.
AI and ML are fundamentally different and do not readily replace DevOps jobs, they just create new types of jobs and improve user experience and capacity to scale. Rob Hesketh @BroadlightInfo Click To Tweet
DevOps will end long term as the roles of dev, test and DevOps converge further.
This will likely happen in 5-7 years, not within the span of an effective AI solution.
ML seems to me unrelated to this discipline except in deployment and training, which a DevOps engineer is likely to cross train into.”
Rob suggests that DevOps will remain the same in the near future because many organisations "still have no clue what DevOps is".
Charles Radclyffe, Emerging Technology Innovation Specialist, Forbes Contributing writer and TEDx speaker
Charles is an innovation specialist, Forbes Contributing writer and TEDx speaker, who has built and sold three tech companies.
He advises organisations on how to develop a strategy for the ethical implementation of AI, Automation and Robotics as well as speaking at events on these subjects.
He also writes a blog on the ethics and societal impact of emerging technology.
He is currently the AI Programme Director for Fidelity International.
Charles believes that AI will transform DevOps in stages, stating,
"Automation of Testing and Monitoring is going to be the first area that is hit, and then infrastructure configuration and management will follow.
DevOps itself is already a fundamental shift in working for many Enterprises, I’m not sure they need AI to disrupt them further just yet!
In the long term, perhaps everything will be automated, but that is really a long way off.
There are some roles that might not exist in the next 10-20 years such as testing or infrastructure management, but full automation is still a pipe-dream.
There are some roles that might not exist in the next 10-20 years such as testing or infrastructure management, but full automation is still a pipe-dream. Charles Radclyffe @dataphilosopher Click To Tweet
Coders are safe (for now)!"
Charles also suggests that,
"Less testers will mean more coders.
Big projects such as Aircraft design are testing constrained, not code constrained.
The Airbus A380 has something like 60 million lines of code, compared to an earlier plane such as the A320 which had only 15 million lines of code.
It’s 4x code, but more like 4-squared times more testing!"
Simon Wardley, adviser to the Leading Edge Forum, where he advises Government and Global Corporations
Simon is an adviser to the Leading Edge Forum, where he advises Government and Global Corporations.
He is a specialist in situational awareness, doctrine and gameplay.
He is an excellent speaker and writer, including his fantastic book, found here https://medium.com/wardleymaps/.
Simon summarised nicely, saying that,
"The future is conversational programming ... think Lambda, think ML, think Alexa ... with Lambda providing the training set, the graph data of function calls not just I/O of functions.
As for DevOps, Cloud (as in IaaS) and containers ... ask me in a decade, I'll probably go ... 'Oh, I remember those'."
Amir Hashmi, CEO zsah, experienced technology, strategy and business process expert
Our own CEO, Amir, is an experienced technology, strategy and business process expert with significant exposure to both the Enterprise and SME spaces.
He founded zsah in 2002 in order to deliver innovative hosted software projects.
zsah now provides organisations of all sizes with private and public cloud services, network and hosted infrastructure, enhanced with advisory and consultative support. zsah is the backbone for many SaaS organisations.
"The focus of DevOps is concentrated on the correct planning and setup of an organisations application operations (and to carry out those operations).
Once set up well, a good DevOps function responds to and predicts changes that are required.
Both of these are ideal for AI and ML to deal with.
Machine Learning works best with rich data sets for it to digest so the value of ML comes with (and increases) with time.
The teams here at zsah are automating DevOps with AI and ML already.
AI and ML are capable of driving huge automation in DevOps and will continue to do so.
DevOps roles will need to change to include other sysadmin or developer/architect roles splitting into ops and project work.
The nature of applications are that they are always changing – sometimes fundamentally – and this means that ML can be set back as it has to learn again or use a limited data set.
AI also then needs re-engineering.
These changes will need human intervention because most organisations will not have dedicated teams of AI and ML specialists dedicated to DevOps.
They will want to buy in software that does that.
The economies involved are interesting as a gulf could grow between smaller organisations and startups using very standard AI and ML tools to automate; medium-sized organisations using little as they require more customisations but do not want to invest in AI and ML teams; and, larger organisations which have teams of AI and ML professionals to carry out in-depth customisations.”
"... DevOps roles are already changing and taking advantage of AI and ML.
The skill sets required for DevOps will increasingly require professionals in this domain to understand and administer AI and ML within the context of their operations."
So that’s what our experts think.
Now it’s over to you, what do you think?
Leave a message in the comments below with your viewpoint on whether AI/ML will enhance or make DevOps roles redundant in the future?
Since we have your attention for a moment, please check out our Gridz solution for private, public and multi-cloud management.