DBR = Death by Robot, but not physical death…just economic death.

I found this website called Replaced by Robot (Will Robots Take My Job? Will My Job Be Automated? · Replaced By Robots!?) a while ago and it really made me think. What jobs will be replaced by automation? Are we heading for an employment cliff? (spoiler alert: no, we are not)

The website was interesting and provided the research that drove the answers. How did Replaced by Robot know an Accountant was 94% likely to be automated? The data is based on a well known University of Oxford study (The Future of Employment: How susceptible are… | Oxford Martin School) which examines how susceptible a job is to automation.

The research data sparked an idea: I work with people to automate tasks all the time. I bet I can look at a list of tasks for a job and understand what can be automated, the right tool for automation and extrapolate how to avoid being automated away.

This experiment led to some interesting revelations. More on that later.

I collected job data from the U.S. Department of Labor’s O*NET system, then used the research driving Replaced by Robot to ascertain the “Likelihood of Automation” for a ONET occupation based on Tasks not the overall job.

Honestly, I don’t think entire jobs will be wholesale automated away instantly…but tasks will. Knowing where to focus your energy as an employee will not only make you more relevant in the AI economy but will help you own your career destiny. And who doesn’t want control over their destiny?

Tasks by Type

I liked how the Oxford paper explored tasks by Labor Type and I kept the same idea. What kind of work is the task (not the job)? Is the task about thinking? Is the task about moving things? I made slight adjustments to account for Perception as a social activity and Manipulation as a separate physical labor type. Here’s the list:

• Cognitive Intelligence: Task takes creative thinking, processing data, analyzing data/processes

• Social Intelligence: Task takes social interaction such as negotiation, collaboration, team work.

• Perception: Using senses to understand what’s happening (yes, technically everything uses your perception but I’m being more direct here in that you need to observe and extract meaning)

• Manipulation: Tasks that take physical environment manipulation like moving things to be in place for someone / something else to work with the thing.

Using this information, I wanted to see how much of a job was Cognitive vs. Social, etc… Knowing that Social activities usually involve “soft skills”, I thought this was a good place to examine closer. Identifying the labor type of tasks will allow me to say “focus here”, “automate here”, “grow here”, which is the conversation necessary to take control of your career destiny.

What did i find?

The future of my past

The first job I reviewed was one I know well from my past: Computer & Information Systems Manager (OSOC CODE: 11-3021.00). I was a software development manager for many years. Here’s what I found when I asked: can this task be automated? If so, how?

According to O*NET, there are 25 tasks that comprise this job. My analysis found that 36% of those tasks can have some level of automation (some tasks can be fully automated). To be clear, I’m not saying 9 tasks can be completely automated. I’m saying that 36% of the tasks can be automated as in all 25 tasks still exists but on average slightly over a third of each task can be automated.

Let’s start with the types of tasks or as we call them above “Labor Types”:

No surprise, most of the work for Computer & Information Systems Manager is “thinking”. This is a danger zone in the world of automation. Thinking can be automated depending on what you’re thinking about. Here are two tasks that are cognitive intelligence (aka thinking) tasks for this job:

1. Analyze data to determine project feasibility

2. Analyze data to inform operational decisions or activities

The first item is a good example of “this is not automatable” because the factors that go into the decision of project feasibility are frequently in flux. Remember, good automation candidates are highly repeatable, high volume, high value. This task is high volume (done often), high value (could impact the business in big ways) but is highly variable depending on the business climate.

The second item is much more standardized. While the decisions might not be automated, the data to drive those decisions will be (or at least mostly automated). Status reports (that’s what this task is about) are high volume (you do it for every project on a regular basis), high value (they tell everyone who needs to know what’s going on) and are very repeatable (if you have a standardized template). Also, chances are if you’re using a tool like Jira, Azure DevOps, Monday.com or whatever, these reports can be generated for you easily.

After analysis, there were 9 out of 25 tasks that were highly susceptible to automation. Those tasks are things like:

• Purchase materials & equipment

• Prepare status reports

• Recruit personnel

This doesn’t mean that 100% of that task is automated, just that more than 1/2 of that task could be automated. Recruiting is a great example. A major part of recruiting is finding and filtering talent. Some of that (NOT ALL - never let automation sort all your candidates b/c you’ll miss some diamonds in the rough) can be done through automation (see Will Interactive Tech and 'Bots' Transform How We Manage Comp and Benefits? (shrm.org)). While automations are handling the grunt work, you can be engaging the community to find your next hire before they get snapped up by someone else.

How would the work be automated?

Using my experience in automation, the stories I hear, the technologies I see, I group tasks by the tool that is best suited for the work.

Tools to automate tasks:

• Digital Assistants: Invisible agents that interact with data or knowledge on your behalf

• Digital Workers: Visible agents that work on your screen to perform tasks (think Robotic Process Automation [RPA])

• Others: There are more but they are not relevant here.

Notice, the Digital Assistant is a future BFF of this job. Tools to automatically build and deliver reports, tools to automatically do code reviews, tools to progress projects through stage gates. Power Automate, Zapier, IFTTT, Alexa, Cortana can all be used to help automate tasks. Here is a list of out of the box integrations available for Azure DevOps through Power Automate (Azure DevOps | Microsoft Power Automate) and this is just the easy stuff you can snap into your automation.

Tools like Digital Workers aren’t a good fit. There isn’t a lot of repeatable data entry or data manipulation for this job so tools like RPA, don’t really have a place today. The capabilities of RPA are ever increasing and perhaps in the near future there will be more of a role but for now, Digital Workers are not a critical tool for this job.

Much of the work is human only. Human Only means that there isn’t a clear “automation” method. Some example tasks include:

• Confer with organizational members to accomplish work activities

• Develop computer or information systems

• Develop operating strategies, plans, or procedures

Can parts of these things be automated? Maybe, but not to a meaningful or valuable measure. Keep in mind, just because something “can” be automated doesn’t mean it should be. The value of automation needs to exceed the cost of automating and some tasks just don’t make that cut.

Now what?

The Verdict

This job will partner closely with Digital Assistants to help with communicating and measuring status. Someone to serve you reports, status, warning signs, and keep the administrative work flowing.

People in this role need to focus on their communication skills and collaboration/negotiating skills to really amplify their impact in the age of Artificial Intelligence. Building consensus, negotiating requirements and project needs, coordinate stakeholders, all of these things are critical to the job’s success and will be the core activities to add even more value to the business.

To get started with collaboration and soft skills, check out Amy Blaschka on Twitter & Forbes.com. She posts lots of helpful soft skill content. Find her here: Amy Blaschka 📲👻✍🏼 (@amyblaschka) / Twitter.

Do This Monday

• Find out what reports can come out of your project systems. Can you use them to communicate status and progress?

• Use the tools you have (Jira/Azure DevOps/Power Automate/Zapier/IFTTT) to get notifications on events that matter for your team (Code Check-in, pull requests, etc…) and use those notifications as an opportunity to connect with your team.

  • Use this connection point to discover the data that isn’t in the data. Reporting systems have lots of information but when you talk to people, listen to their stories, you can find information that isn’t in your project status system. Use your social intelligence to get the whole story so you can improve the project status reports with content not found in the data.

• Identify your customers and connect with them to build social report. Build an empathy map to understand them better (how do I do an empathy map?).

• Build your organization’s talent brand by getting into the community where your potential hires dwell (think Stack Overflow or LinkedIn Groups) and make connections.

What do you think?

I’ve built out tools to explore more occupations. Would you like to see more? Your feedback will tell me to keep going or nah :)

Let me know!

Thanks for reading!

Header photo from Stack of magazines & pencils in gray cup · Free Stock Photo (pexels.com)

Last thing…

What is an Empathy Map? Here I talk about empathy maps in context of writing fiction but you can extrapolate how to use this for assessing users, customers, partners, etc… Enjoy!