As I mentioned in last month's column ("Mobile App Madness," article ID 65317), my son recently graduated from college and landed a job in the software industry. In his pursuit of that job, he worked with a number of placement agencies. To keep his life organized, he used a feature on his Google Nexus One phone that transcribed voice messages into email. This allowed him to focus all his attention on one source of information—email.

The automated transcription of voice to text has certainly improved over the years. But it's pretty clear that we haven't completely mastered this technology. As evidence of this, I submit the following transcription my son received in email:

Hi, I'm calling about a job opportunity I currently have open. I didn't know if you are still in the job market. I'd be very interested in letting you know more about it and I love you. Please give me a call back. My direct line is . . .

Now, my son is a lovable guy, but I seriously doubt that's what the placement agent really said. When I read the email, I had two reactions. My first was to laugh hysterically. My second was to question our relentless pursuit of automation in the IT industry. This led me to wonder if sometimes we push automation too far.

Hot: Automation As a Journey

Automation has several excellent use cases in the IT industry. One such use case is to automate repetitious manual tasks. There's no point in an operator or administrator doing the same thing over and over and over again if the task can be automated. A good example of this is the automation of employee onboarding: Automating all the steps for account provisioning and asset procurement makes sense because the process is well known, predictable, and repeatable.

Another common use case for automation is to bridge disparate technologies. Despite the best efforts of many vendors, most organizations buy hardware and software from multiple sources. Therefore, we often find ourselves having to integrate information from multiple sources. A good example of this use case is using automation to take data gathered from a network discovery tool and integrate it into a configuration management database (CMDB).

In both these cases, the key to success is a focus on a specific problem or a well-defined set of tasks. Focused automation implementations are "hot" in my eyes.

Not Hot: Automation As a Destination

If focused automation is one end of the spectrum, the other end is the concept of the fully automated data center. I'm sure you've heard the pitch—it's quite the rage in today's virtualized, cloud-hugging IT market.

The vision is a data center that is self-tuning based on the workloads it's running and demand generation. If end-user demand increases on a workload, the workload is relocated to a higher-performing resource pool (in the data center or in the cloud) to handle that demand. Or if a rack of computers exceeds energy or cooling policies, the workload is moved off the rack to keep the data center in compliance with energy guidelines. In this vision, everything is automated. Policies dictate what happens and under what circumstances.

I must confess that for a long time I was supportive of this concept. Then I started to wonder what would happen five years after we implement completely automated data centers and assign our IT personnel to other tasks (e.g., business alignment, relationship management). What if something really goes wrong? Will we have enough knowledge of the current workloads, resource allocations, demand parameters, and resource requirements to manually correct serious problems? Or will we find ourselves in a variation of the 1973 Westworld movie, in which the androids start killing the humans, and no one can figure out why they're doing it or how to stop them?

Call me paranoid if you like, but the concept of automation growing beyond our human capacity to comprehend the details is definitely "not hot" in my eyes.

Sean Chandler is a computer and network consultant with more than 30 years of field experience. Astro, a border collie with more than 40 dog years of data processing experience, provides technical support to his master, Sean.

Astro's Pick of the Litter

You'd think my master would know better. He's experienced many disastrous in-place upgrades of Windows clients and servers over the years. He even had major trauma with an upgrade from Leopard to Snow Leopard on his precious black MacBook (for a description of that adventure, see "Old Home Computers: Reuse, Recycle, Regret," January 2010, article ID 64543). But for some reason, he didn't think twice about upgrading his second-generation iPhone to the latest operating system release, OS 4. Big mistake. Although his phone is technically functional after the upgrade, it offers the performance of a brick. He's so very, very disappointed—mostly with himself. On the other hand, that didn't stop him from ordering a new iPhone 4 to address the situation!

—Astro