Read the Job Market Before It Changes

How AI Is Reshaping Every Profession

Part 1: The Shift Is Already Happening. Whether You Planned For It or Not

Here's what those headlines mostly missed: Klarna didn't just fire 700 people and call it a day. The company simultaneously reported that its remaining human agents were handling more complex, higher-value escalations, the conversations where empathy, judgment, and accountability actually matter. The AI took the volume. The humans took the weight. That's a meaningful distinction, and it plays out across nearly every profession right now. The question isn't whether AI is changing your field. It's changing what parts of your job look like, which skills become more valuable, and what you need to do differently starting now.

The Klarna story isn't a customer service story. It's a template. Across law firms, marketing agencies, hospitals, schools, consulting practices, and HR departments, the same basic pattern is emerging: AI absorbs the high-volume, repeatable work, and the humans who remain, or thrive, are the ones who understand how to work alongside it. The professionals who are struggling aren't the ones whose jobs disappeared. They're the ones who didn't notice the job was changing until it already had.

What 'AI Reshaping a Profession' Actually Looks Like Up Close

Consider what happened at the law firm Allen & Overy, one of the world's largest, when they rolled out Harvey, an AI legal assistant built on GPT-4, to their lawyers in 2023. Before Harvey, a junior associate might spend four hours reviewing a contract for standard risk clauses, producing a first-draft memo for a senior partner to review. With Harvey, that same first-draft memo takes about 20 minutes. The associate still reviews it, edits it, applies their legal judgment, and owns it professionally. But they can now handle three times the volume in a week, and the senior partners are getting better-prepared briefs because the AI never forgets to check a clause it's been trained to look for.

What changed for those lawyers isn't their expertise. It's their workflow. The junior associate who resisted Harvey, or didn't bother learning it, now looks slower and more expensive than the one who adopted it. This isn't a story about AI replacing lawyers. It's about AI creating a visible gap between lawyers who adapted and lawyers who didn't. The same gap is opening in marketing teams, finance departments, HR functions, and classrooms right now. In most cases, the gap isn't enormous yet. But it widens every quarter as the tools improve and adoption spreads.

The principle here is simple but easy to underestimate: when AI handles the time-intensive, repeatable parts of a job, the professionals who thrive are the ones who use that freed-up capacity for higher-value work, not the ones who simply do less. At Allen & Overy, the associates using Harvey aren't working 20-hour weeks. They're taking on more clients, doing deeper analyzis, building stronger relationships with partners, and developing expertise faster because they're seeing more varied work. The tool didn't shrink their role. It compressed the parts that didn't require them, so the parts that do require them expanded.

A Different Industry, The Same Pattern: Marketing at Publicis

Publicis Groupe, one of the world's largest advertising holding companies, launched an internal AI platform called Marcel in 2019, and by 2023 had deeply integrated generative AI tools across its creative and account teams. Their copywriters use AI to generate first-draft ad copy across multiple variations simultaneously. A campaign brief that once produced 3-4 headline options after a half-day brainstorm now produces 40 variations in under an hour. Account managers use AI to draft client status reports, competitive analyzes, and campaign performance summaries. The human work is the strategy, the taste, the client relationship, and the final edit. The AI does the scaffolding.

What's striking about Publicis isn't the technology, it's the organizational expectation that shifted. By 2023, fluency with AI tools became part of how Publicis evaluated and promoted talent. A creative director who couldn't articulate how they were using AI in their workflow was seen as behind the curve, not principled. That's a cultural shift, and it's spreading across industries faster than most professionals realize. The agencies, firms, and companies that move first on AI adoption tend to set the pace for what 'normal' looks like in their sector, and their competitors follow within 12 to 24 months.

What This Looks Like for a High School Principal in Nebraska

The Klarna and Allen & Overy stories involve large organizations with dedicated technology budgets. But the same shift is landing in far less glamorous settings. Take a high school principal in a mid-sized school district, the kind of administrator who spends 40% of their week on documentation: drafting staff communication emails, writing policy updates, preparing board meeting summaries, creating professional development agendas, and responding to parent concerns in writing. None of that work requires a principal's professional expertise. It requires time and clear writing. A principal using ChatGPT Plus to produce first drafts of all those documents can reclaim 8 to 12 hours a week, time that goes back to being in classrooms, coaching teachers, and building the school culture that no AI can build.

Several school districts have started reporting exactly this outcome. Administrators who adopted AI writing tools in 2023 and 2024 describe the experience as getting a highly capable administrative assistant, one that's available at 10pm, never complains, and produces a clean first draft in 90 seconds. The principal still owns every word that goes out under their name. They edit, they adjust the tone, they add the context that only they have. But the blank-page problem, the friction of starting a complex document, largely disappears. This frees the cognitive and emotional bandwidth for the work that actually requires a skilled, experienced human being in the room.

Why This Moment Is Different From Previous Technology Shifts

Every generation of professionals has lived through technology shifts that changed how work gets done. Spreadsheets replaced manual bookkeeping calculations. Email replaced memos and faxes. Search engines replaced research librarians for routine information retrieval. Each shift created anxiety, displaced some roles, and elevated others. What's different about AI tools in 2024 and beyond isn't the fact of change, it's the scope and speed of it. Previous tools automated specific, narrow tasks. AI tools operate across language, analyzis, summarization, ideation, and drafting simultaneously. A single tool like Claude Pro or ChatGPT Plus can help you write, research, plan, critique, and communicate, all in one session.

The speed matters too. When spreadsheets arrived, it took roughly a decade for adoption to become standard across industries. When email arrived, the transition took five to seven years to feel universal in office environments. Generative AI tools went from niche to mainstream professional conversation in under 18 months. ChatGPT reached 100 million users in two months, the fastest consumer technology adoption in history at that point. By 2024, Microsoft had embedded Copilot directly into Word, Excel, Outlook, and Teams, tools that hundreds of millions of professionals already use every day. The learning curve for access is essentially gone. What remains is the learning curve for effective use.

This speed changes the professional calculus. With previous technology shifts, a professional could wait and observe before adopting, and the gap between early adopters and late adopters was manageable. With AI tools spreading this fast, the gap between professionals who are building fluency now and those who are waiting opens quickly. A salesperson who uses Microsoft Copilot to prep for every client call, draft follow-up emails, and summarize CRM notes is compounding an advantage every single week over a peer who isn't. After a year, that compounded difference shows up in pipeline metrics, quota attainment, and performance reviews, whether the manager explicitly connects it to AI use or not.

Click to enlarge

Key Principles From Part 1

1. AI restructures professions, it doesn't simply eliminate them. The Klarna, Allen & Overy, and Publicis examples all show the same pattern: AI absorbs high-volume, repeatable work while humans retain the judgment-intensive, relationship-driven, and accountable work.
2. The professionals who struggle most aren't those whose jobs disappear, they're the ones who don't notice the job is changing until the gap between them and their AI-fluent peers becomes visible in performance metrics.
3. No role is too small or too 'human' to be affected. A school principal's administrative workload is just as restructurable as a law firm associate's contract review process. The tools are the same; the workflows are different.
4. The speed of AI adoption is historically unusual. Tools that previously took a decade to become standard are embedding into mainstream professional software. Word, Outlook, Teams, Salesforce, within months. Waiting to observe is a faster-closing window than in previous technology shifts.
5. Your first practical move is a workflow audit: identify which parts of your job require your specific expertise and which parts mostly require time. The second category is where AI delivers immediate, measurable value.
6. Fluency with AI tools is already becoming a professional differentiator. At firms like Publicis and Allen & Overy, it's part of how talent is evaluated. That expectation will spread to most industries within the next two to three years.

When AI Meets the Messiest Jobs: The Human Services Paradox

In 2023, Unilever's HR team faced a problem that spreadsheets couldn't solve. They were screening over 1.8 million job applications annually across 190 countries. Recruiters were burning out. Good candidates were slipping through. Hiring timelines stretched to 90 days or more. Their solution was to deploy AI screening tools that could parse resumes, flag skill matches, and rank applicants before a human ever opened a single file. The result: hiring cycle time dropped by 75%, and recruiter hours spent on initial screening fell dramatically. On paper, it looked like a clean AI win.

But here's what the headline numbers missed. Unilever's recruiters didn't disappear. They shifted. Instead of reading the 800th resume of the week with glazed eyes, they were spending their time on video interviews, candidate experience design, and salary negotiation, the moments where human judgment actually matters. The AI handled volume. The humans handled nuance. One senior recruiter described it this way: 'I used to feel like a sorting machine. Now I feel like a talent advisor.' That shift in self-description tells you everything about what AI actually does to professional roles when it's implemented thoughtfully.

The principle here isn't that AI eliminates jobs. The principle is that AI eliminates tasks, and when it eliminates enough tasks from a role, the role itself transforms. The recruiter didn't become obsolete. The role of recruiter quietly evolved into something more strategic, more relational, and more valued. This is the pattern playing out across industries right now, and it's happening faster than most professionals realize. The question isn't whether your role will be touched by AI. It's whether you'll shape that transformation or just experience it.

The Sales Rep Who Stopped Cold Calling, and Started Closing More

Marcus runs a six-person B2B sales team at a mid-size logistics company in Atlanta. Eighteen months ago, his reps spent roughly three hours a day on what he calls 'prep theater', researching prospects on LinkedIn, writing personalized outreach emails, updating CRM notes after calls, and building pipeline reports for Monday morning meetings. The actual selling, listening, handling objections, building trust, got maybe four hours of the workday if they were lucky. Then Marcus started using a combination of ChatGPT Plus for email drafting and HubSpot's AI tools for CRM automation.

Within two months, his reps reclaimed nearly two hours of daily selling time. Cold outreach email quality went up because they were using AI to personalize at scale rather than copying generic templates. Post-call notes wrote themselves. Pipeline reports generated automatically. One rep, who had been quietly considering leaving because she felt 'like an admin with a quota,' became the team's top performer in Q3. The tasks AI absorbed were the ones she hated. The work that remained, reading a room on a discovery call, knowing when to push and when to listen, was exactly why she got into sales.

This story surfaces a principle that goes beyond sales: AI's impact on a role depends enormously on which tasks it absorbs. When AI takes the repetitive and the routine, professionals often report higher job satisfaction alongside productivity gains. When AI is deployed poorly, replacing judgment calls with automated decisions, removing the human from sensitive conversations, the results are very different. The technology is the same. The implementation is everything. Marcus understood this intuitively. He didn't automate his team's relationship-building. He automated their paperwork.

Healthcare's Uncomfortable Reckoning

The healthcare sector is where the AI transformation gets genuinely complicated. At Intermountain Health, one of the largest hospital systems in the American West, AI tools are now used to flag patients at risk of sepsis, predict readmission likelihood, and help radiologists prioritize which scans to review first. In some diagnostic tasks, AI models match or exceed average radiologist accuracy on specific conditions like detecting diabetic retinopathy. These are not hypothetical capabilities. They are deployed systems changing how care is delivered today.

But ask any nurse practitioner or attending physician what this has actually changed about their day, and the answer is almost never 'I do less work.' It's usually the opposite. AI surfaces more information, faster, which means clinicians make more decisions per shift, not fewer. The cognitive load shifts from 'finding the signal' to 'interpreting the signal and acting on it.' A radiologist who once reviewed 80 scans per day might now review 110, with AI pre-sorting and flagging the high-priority ones. Productivity rises. Burnout risk, if the transition isn't managed carefully, can rise alongside it. The technology changes the work. It doesn't automatically make the work easier.

How AI Is Reshaping Roles Across Industries

Click to enlarge

The Teacher Who Stopped Dreading Sunday Nights

Sarah teaches 10th grade English in a public school district outside Chicago. Like most teachers, she was spending 8-10 hours every weekend on lesson planning, rubric creation, progress report writing, and designing quiz questions, work that had nothing to do with actually being in front of her students. Last year, she started using Claude Pro to generate first drafts of lesson plans aligned to her state standards, build differentiated reading comprehension questions at three difficulty levels, and draft the narrative sections of student progress reports. Her estimate: she reclaimed four to five hours every weekend.

What did she do with those hours? She spent more time reading student essays carefully. She started a weekly one-on-one check-in with struggling students that she'd never had bandwidth for before. Her students' writing scores improved measurably over the semester. Sarah is careful to say the AI drafts were never used as-is, she edited every lesson plan, rewrote most of the report language to reflect what she actually knew about each child. But the blank page problem was gone. Starting from something coherent and editable is categorically different from starting from nothing at 9pm on a Sunday. That difference, multiplied across a school year, is significant.

What This Means for How You Think About Your Own Role

Look at your own job for a moment. Not your job title, your actual job. The list of tasks you did last week. If you're honest, a significant portion of that list is probably things like: drafting routine communications, compiling information from multiple sources, creating documents you've created dozens of times before, preparing summaries of things you already understand. These are exactly the tasks AI handles well right now. They're not unimportant tasks, they take real time and effort. But they're not the tasks where your specific expertise, judgment, and relationships create the most value.

The professionals who are thriving in AI-augmented workplaces share a habit: they've gotten brutally honest about which parts of their work are high-value and which parts are high-volume. High-value work is where your judgment, relationships, context, and expertise produce outcomes no AI can replicate. High-volume work is where you're primarily executing well-defined, repeatable processes. AI tools are excellent at high-volume. They're genuinely poor at high-value, at least for now, and especially in the specific context of your organization, your clients, and your industry knowledge. Your job is to direct AI toward volume so you can protect time for value.

This isn't a passive realization. It requires active decisions about how you work. It means experimenting with tools like ChatGPT Plus, Microsoft Copilot, or Notion AI on real tasks this week, not waiting for your company to roll out a formal AI training program. Professionals who develop hands-on fluency now, while AI adoption is still uneven, build an advantage that compounds over time. The people who will be most valuable in three years aren't necessarily the ones with the most AI knowledge. They're the ones who combined AI fluency with deep domain expertise, and used each to amplify the other.

Click to enlarge

Key Lessons from Part 2

1. AI automates tasks, not jobs. Most roles contain a mix of automatable and non-automatable work, the proportion matters more than the job title.
2. When AI absorbs routine tasks, the remaining work typically shifts upward in complexity and strategic value, which can increase both performance and job satisfaction.
3. The same AI tool can help or harm a workforce depending entirely on implementation. Automating paperwork is different from automating judgment.
4. The 'first draft' habit is the most accessible entry point for non-technical professionals, it eliminates the blank page problem across almost every work output.
5. Healthcare shows that AI increasing diagnostic support doesn't reduce clinician workload automatically, it changes the nature of cognitive demands without reducing them.
6. Professionals who combine AI fluency with deep domain expertise outperform those who develop either skill in isolation.
7. Your personal role audit, separating high-volume from high-value tasks, is more useful than any generic industry analyzis of AI's impact on your profession.

The Human Edge in an AI-Augmented World

In 2023, Klarna, the Swedish fintech company, made headlines by announcing its AI assistant was handling two-thirds of customer service chats, doing the work of 700 human agents. Leadership celebrated the efficiency. Then, quietly, something shifted. Customer satisfaction scores dipped. Escalations increased. The AI was fast, accurate on policy, and tireless, but it couldn't read frustration, defuse anger, or recognize when a customer needed to feel genuinely heard rather than efficiently processed. Klarna didn't abandon the AI. They restructured the team so human agents handled the emotionally charged cases while AI handled routine ones. The lesson wasn't that AI failed. It was that the humans who understood how to work alongside it became dramatically more valuable.

This story plays out across industries in different costumes. A law firm replaces junior associates' document review hours with AI, then realizes the partners who can interrogate AI output, spot subtle errors, and apply strategic judgment are worth more than ever. A school district rolls out AI tutoring tools, then discovers that teachers who coach students on how to learn with AI produce better outcomes than those who either ban it or step back entirely. The pattern is consistent: AI compresses the routine, and the professionals who thrive are the ones who own what remains.

The core principle here isn't motivational, it's structural. When a tool automates the bottom layer of a job, the middle layer becomes the new bottom, and the top layer becomes the new competitive advantage. Radiologists who once spent 60% of their time reviewing normal scans now spend that time on ambiguous cases, second opinions, and patient consultation. The job didn't disappear. Its center of gravity moved upward. Professionals who recognize this shift early and deliberately build skills in judgment, communication, and AI oversight will not be replaced. They will be the ones doing the replacing.

Three Professions, Three Transformations

Consider what happened at JPMorgan Chase when they deployed COIN, a machine learning program that reviews commercial loan agreements. The task previously consumed 360,000 hours of lawyer and loan officer time annually. COIN does it in seconds. The lawyers didn't disappear. Their job description changed overnight. Instead of reading contracts line by line, they now review AI-flagged anomalies, advise on edge cases the model hasn't seen before, and focus on the negotiation strategy that no algorithm can replicate. The volume of work they can handle tripled. The nature of the work shifted entirely toward high-stakes judgment.

In a completely different sector, the marketing team at Heinz ran an experiment: they asked several AI image generators to produce 'ketchup.' Every image, unprompted, produced something that looked unmistakably like Heinz. The brand team used this as strategic insight, proof of brand dominance so strong it had shaped AI training data. They built a global campaign around it. The campaign won awards. No AI conceived the strategic insight, commissioned the experiment, or decided it was worth celebrating publicly. A human marketer connected dots across brand strategy, consumer psychology, and cultural timing. The AI was a tool in that story, not the author of it.

Then there's teaching. When Khan Academy launched Khanmigo, an AI tutor built on GPT-4, it could answer student questions at 3am, provide infinite patience, and adapt explanations in real time. Teachers initially felt threatened. The ones who leaned in discovered something different: they now had data on exactly where each student was struggling, before class, every day. They stopped spending time re-explaining concepts the AI had already covered and started spending it on discussion, mentorship, and the social-emotional work that no chatbot can do. Their effectiveness per hour of teaching time increased measurably.

Building Your AI-Augmented Practice

The professionals gaining ground right now aren't the most technical ones, they're the most deliberate ones. They've mapped their own jobs into two columns: tasks that are essentially information processing (drafting, summarizing, researching, formatting, scheduling) and tasks that require human presence (negotiating, deciding, empathizing, leading, creating meaning). They're systematically shifting their time toward the second column by using AI tools for the first. This isn't a productivity hack. It's a career repositioning strategy executed one workflow at a time.

Getting good at working with AI tools requires a skill that has no technical name but a clear business one: knowing how to brief a collaborator well. When you give ChatGPT a vague instruction, you get a vague result, exactly like a new hire who wasn't properly onboarded. When you give it context, constraints, a target audience, and a specific output format, you get something genuinely useful. This is prompt craft, and it is learnable in hours, not months. The professionals who master it first in their team or department will have an outsized advantage for the next several years.

The final piece is posture. AI tools will keep improving. A tool that produces mediocre output today may be excellent in 18 months. Professionals who stay curious, who periodically re-test tools on tasks they previously dismissed, will keep finding new efficiencies. Those who made one judgment call in 2023 and never revisited it will fall behind, not because AI replaced them, but because their colleagues kept adapting and they didn't. Staying current doesn't require technical expertise. It requires the same habit you'd apply to any professional skill: regular, intentional practice.

Click to enlarge

Key Takeaways

1. AI compresses routine work, the professionals who thrive redirect their time toward the judgment, relationship, and strategic work that remains.
2. No profession is being eliminated wholesale, but every profession is being restructured. The center of gravity in each role is shifting upward toward higher-complexity tasks.
3. Real-world examples from law, finance, education, and marketing show a consistent pattern: AI handles volume and speed; humans handle ambiguity, trust, and meaning.
4. The skill that matters most right now isn't coding, it's knowing how to brief AI tools clearly and evaluate their output critically.
5. Professionals who test AI tools regularly and update their workflows accordingly will outpace those who made a single judgment call and stopped experimenting.
6. Your competitive advantage lives in the tasks AI consistently underperforms: ethical judgment, emotional intelligence, contextual strategy, and authentic relationship-building.
7. Starting small and specific, one task, one tool, one workflow, is more effective than trying to overhaul everything at once.