The Human Job in the Machine Age

Today’s class made the future feel less like a sci‑fi trailer and more like a calendar notification. AI is no longer a distant robot arm in a factory. It is a tab open beside our homework, a quiet second brain in the office, and maybe — this is the uncomfortable part — the intern who never sleeps.

“AI is not a humiliation for humanity. Maybe the humiliation is thinking that being the smartest thing was all we had.”Timothy Chung, AI & Work class discussion — reconstructed from student notes

For Management Analysts, the story is not “safe” or “doomed.” It is stranger. Will Robots Take My Job lists the role at 19% calculated automation risk, with users voting a higher 37% chance of full automation within twenty years. That gap feels honest: the model says the job survives; people’s stomachs say, “Are we sure?”

The job is not disappearing. The job interview is changing.

“People said painting was dead after photography. Instead, art found new approaches. AI may do the same to work.”Timothy Chung, class notes, referencing Paul Delaroche

A Management Analyst studies how organizations work, finds inefficiencies, designs better procedures, and advises managers. AI can already chew through meeting notes, compare costs, draft slides, summarize reports, and spot patterns faster than a tired student on coffee number three. That is not a small threat. That is half the desk.

But the other half of the desk is human: convincing a defensive manager that their process is broken, noticing when a team says “fine” but means “furious,” and translating a spreadsheet into a decision people will actually follow. In class, I kept thinking: AI is excellent at answers, but work is often about getting humans to accept them.

From class: Hummingbird and the feedback problem

Mr. Chung’s example of Hummingbird made AI feel practical, not just futuristic. A system that gives consistent feedback on thousands of essays can do in minutes what humans struggle to do at scale. In our notes, the striking number was an 80% reduction in operational cost while keeping feedback quality consistent.

“Hummingbird gives the feedback professors used to give — faster, consistently, and across thousands of essays humans cannot check as quickly.”Timothy Chung, class discussion — student notes

Human evidence #1: I am writing this with a tiny knot in my chest because “AI can help you work faster” sounds great until you realize employers may hear “fewer beginners needed.”

The broader evidence is mixed. Anthropic’s Economic Index found AI use leaning more toward augmentation than full automation — 57% collaborative use compared with 43% delegated automation. McKinsey’s 2025 survey says 88% of organizations now use AI in at least one business function, but many are still experimenting instead of transforming everything overnight.

One chart: the pressure on Management Analysts

Financial Times Visual Vocabulary choice: bar chart, comparing job risk and labor-market signals from the same occupation profile.

The “job score x10” bar converts Will Robots Take My Job’s 8.1/10 score into 81% only to fit the same 0–100 scale. It is not an automation probability.

Human strengths

Management Analysts survive by doing the parts of work that happen between the data points.

Creative thinkingProblem solvingPersuasionSocial perceptivenessAdvising managers

AI threats

AI will attack the repeatable tasks first: research summaries, cost comparisons, process maps, slide drafts, meeting notes, and “first version” recommendations.

Class note: “AI can hallucinate, and it can’t feel human emotions.” — Timothy Chung

Data analysisReport draftingWorkflow documentationBenchmarkingJunior analyst tasks

McKinsey’s report made the office future feel less like a clean replacement and more like a redesign. Nearly two-thirds of organizations had not yet scaled AI across the enterprise, even while 62% were at least experimenting with AI agents. Translation: the revolution is real, but it is also stuck in meetings.

Mr. Chung also brought up Creverse and Korean hagwons — learning centers for English, math, and coding — and asked whether AI will kill them or make them disappear in twenty years. That question belongs in this article because every workplace is a kind of learning center now: people constantly retrain, adapt, and prove they are still useful.

“Will AI kill learning centers? Or will it redefine what they do over the next twenty years?”Timothy Chung, class discussion — student notes

Human evidence #2: After class, I imagined a future manager saying, “The AI made the recommendation,” and everyone in the room pretending that means nobody is responsible.

That is where Management Analysts could matter more, not less. Someone has to ask whether the data is biased, whether the recommendation ignores workers, whether a “productivity gain” is just a nicer phrase for layoffs, and whether the new process makes people feel like replaceable furniture.

One quote from the job’s public discussion hit me because it sounded like a classroom debate: “Management deals with different personalities,” one commenter wrote, arguing that unpredictable human behavior is hard to automate. Another warned that analysts will need to think “beyond the obvious.” Both can be true. AI may steal the obvious. Humans must become better at the non-obvious.

Human evidence #3: Seoul makes this feel sharper. In the subway, everything is efficient; in a classroom, everything is emotional. The future of work is probably both: a perfect system full of imperfect people.

“One person can now do alone in an hour what used to take a team weeks.”Timothy Chung, class discussion — student notes

My hope is that AI becomes a power tool, not a trapdoor. A Management Analyst with AI could test more scenarios, interview more carefully, visualize problems better, and spend less time formatting slides at 1:00 a.m. My fear is that companies use AI to remove the apprenticeship ladder: the boring tasks are often how beginners learn judgment.

So the advice I would give my future self is simple: do not compete with AI at being fast. Compete at being trusted. Learn the tools, yes, but also learn to ask better questions in rooms where people are nervous, proud, or quietly afraid.

Eight pieces of evidence of humanity

I mention today’s class instead of pretending the article came from nowhere.
I integrate Timothy Chung’s class ideas about Hummingbird, hagwons, hallucinations, and painting after photography.
I admit anxiety about AI and future jobs.
I use a Seoul observation: efficient systems, emotional classrooms.
I include a personal fear about “almost as good” replacing beginners.
I connect the topic to student life: coffee, homework, and late slide formatting.
I disagree with a simple doom/safe answer.
I end with advice I would actually give my future self.

Sources used:

Will Robots Take My Job, “Management Analysts,” for automation risk, wages, growth, volume, strengths, and public comments. Anthropic Economic Index for augmentation-versus-automation patterns and occupational AI-use context. McKinsey, “The state of AI in 2025,” for organizational AI adoption, agent experimentation, workforce expectations, and scaling data. Timothy Chung class notes supplied by the student, for lecture-specific quotes and examples.

Will Robots Take My Job · Anthropic Economic Index · McKinsey State of AI 2025 · FT Visual Vocabulary