Back to blog

When should you automate a job instead of assisting it?

Thắng Đoàn
Thắng Đoàn

You have a budget for AI and pressure to use it. The costly mistake is automating work that needed human judgment all along.

Automation and augmentation are both happening

Boston University frames the core leadership decision as deciding what should be automated, what should be augmented, and what should remain under direct human judgment. Harvard Business School's Suraj Srinivasan finds the impact of AI is not uniform across jobs. His research with companies shows automation potential and augmentation potential are both being implemented in substantive ways.

Automate the mind-numbing, not the person

Boston University quotes Slack co-founder Stewart Butterfield: "There's a lot of automation that can happen that isn't a replacement of humans but of mind-numbing behavior." TechTimes reports routine processes like documentation, data entry, and reporting are now managed by AI systems. That is your automation candidate list: repetitive cognitive work with clear rules and low judgment.

  • Documentation, data entry, and reporting: automate, per TechTimes
  • Work needing judgment or discretion: augment instead

Augmentation keeps judgment in the loop

Boston University notes that when people stop performing monotonous tasks, they gain freedom to focus on more challenging work, which is where augmentation comes in. TechTimes calls human-AI complementarity a defining force in modern workplaces, where humans and intelligent systems collaborate to improve accuracy and efficiency. The test is simple: if a wrong output creates real damage, a human stays in the loop.

What this does not solve

None of the sources give a task-by-task scoring rubric for your specific jobs. Srinivasan notes AI also changes how people get hired, not just how work is done. Occupational transitions are still shifting, per TechTimes, so today's automate-or-assist call may be wrong in a year. You still own the decision rights and the accountability.

Which three tasks on your team are truly mind-numbing rather than judgment-based?
Where in your workflows would an AI error cause damage no one would catch?

List five tasks your team repeated last week, then mark each one automate or augment.

Sources

TechTimes: Future of Work

Harvard Business School: How AI is impacting the job market

Boston University: AI Augmentation vs. Automation

Share:

Recommended for you

Enjoyed this article?

Subscribe for new articles. No spam. Unsubscribe anytime.

By subscribing you agree to receive the newsletter. No spam, and you can unsubscribe anytime.