AI cloudops is coming, whether you like it or not

As someone who’s worked with AI for the last 30 years (yes, it was a thing 30 years ago), I’ve often thought of its capabilities were overrated and used for the wrong things in many cases. Now that it’s cheap thanks to cloud computing, and much more effective thanks to the pace of innovation, AI as a solution is coming up again, including the use in cloud operations.

The idea is to replace people with AI to be both proactive and reactive to cloud operational issues such as outages, resource governance, security attacks, and performance. Cloudops involves largely repeatable problems, right?

There are of course some upsides and some downsides to this. Moreover, although the use of AI in cloud operations maybe a foregone conclusion, there will still be a learning curve that is required. As long as you understand that and know what to expect in terms of ROI for both the short term and long term, I’m okay with anything that that makes cloud operations more effective.

So, let’s look at the pros and cons.

The pros of AI for cloudops

The pros are that you can have a 7/24/365 monitoring and management program on the cheap. If you believe operational staff is expensive, try hiring them for shift work. AI-based monitoring and management systems never sleep, never take time off, and never ask for a raise. Once they are up and running, they cost almost nothing beyond their license fees and infrastructure costs. And they are self-learning at the same time; in other words, the more they run, the better that they get at the job.

Another pro is that these systems get smarter every day and share a common brain. People get smarter with experience as well, but they don’t do a good job sharing their experiences with others. People also retire and quit, with the knowledge and experience walking out the door with them.