Ethan Lawson
In July 2025, I hired three junior developers. By October, two of them had been caught in the AI layoff wave—not at my company, but across the industry. 77,999 AI-attributed tech job losses occurred in the first six months of 2025 alone, with software developers in their early twenties getting hit hardest. I watched them panic, pivot, and eventually land somewhere better. But the two-year clock was already ticking. This is the story of what I learned building products in the middle of a job displacement nobody fully prepared for.
The Math Nobody's Talking About Honestly
Let's start with the brutal numbers, then the reason for hope. According to the World Economic Forum's Future of Jobs Report 2025, 92 million roles could be displaced by 2030. That's terrifying. But in the same breath: 170 million new roles are projected to emerge. That's a net gain of 78 million jobs. The asymmetry matters.
Here's what makes this different from previous waves: McKinsey research from 2025 shows that roughly 57% of current U.S. work hours involve tasks AI systems could technically handle. That's not the same as saying 57% of jobs will vanish. It means half the work people do every day could be automated—but automation doesn't delete roles, it transforms them. The catch: companies adopting AI aggressively are making that transformation happen faster than slower adopters, and the gap between those two groups is where careers get made or stalled.
The real issue: that 78 million net gain is distributed unevenly across time, geography, and skill levels. A customer service rep in Ohio whose role is automated doesn't automatically become a prompt engineer in San Francisco. The gap between those two realities is where the genuine human cost lives.
Why Your Current Role Is Either a Launchpad or a Trap
Not all jobs face equal automation pressure. Here's what matters: the type of work you do, and whether your company is racing toward AI adoption or treating it like a distant threat.
If you're in customer service, routine data entry, or basic administrative work, you're in the highest-risk category. If you're managing people who do that work, your automation exposure is significantly lower. This isn't random—it's structural. Management roles require judgment calls that AI still struggles with. Execution roles are exactly what AI was built to handle.
The second variable is your company's AI velocity. MIT research shows that companies aggressively adopting AI experience higher employment growth and significantly higher wage growth compared to peers. That's not speculation—that's showing up in actual hiring and comp decisions right now. The choice isn't "stay or leave." It's "stay in a company racing toward AI adoption or one that's still debating whether it's real."
The Two-Year Window I Didn't Account For
Here's what I got wrong: I thought displacement and opportunity were happening simultaneously. They're not. According to Stanford's 2025 AI Index Report, career transitions take roughly two years. Someone loses a job in 2025. They spend six months upskilling or pivoting. They land in a new role in early 2026. They're fully ramped by late 2026.
That means the critical window is 2025-2030. Decisions made right now determine whether you're positioned to move into the 170 million emerging roles or competing for scraps in the jobs being shed. The two-year lag is the delta between when AI reaches your role and when you're actually employed doing something new. If you wait until your job is at immediate risk, you've already lost two years of positioning time.
This is why I stopped hiring entry-level talent for pure execution roles in 2024. Not because I'm heartless—because I knew those roles were about to accelerate into AI adoption curves. Better to hire for roles where AI amplifies judgment rather than replaces it.
What I Actually Did When the Displacement Started
When two of my hired developers hit the layoff wave, I made three moves with them and with my own positioning:
Move One: Shift output from execution to decisions. Instead of measuring people on "lines of code shipped" or "tickets closed," I started measuring on "decisions that mattered." That moves you from the 67% automatable category toward roles where AI is a tool that amplifies judgment, not a replacement. For those developers, that meant moving from "implement what the PM designed" to "what should we actually build?"
Move Two: Build visibility inside AI-forward companies. The companies growing 6% faster are also paying more. They're also more stable. I helped both developers network into those companies while they were still employed. One landed at a fintech company aggressively building AI infrastructure. The other moved to a mid-stage startup that was organizing around AI-augmented workflows. Neither was waiting for their old job to implode.
Move Three: Learn the one skill your 2027 role will need. For the developers, that was learning how to work with AI agents, not compete with them. For me, it was understanding which roles were going to be amplified by AI and which were going to be hollowed out. I started spending time with operations teams, strategy teams, and people managing AI implementations. That visibility changed everything about how I hire, promote, and structure work.
Which Jobs Survive, Which Don't—And the Honest Timeline
Customer service is the clearest example of high automation risk. AI systems like Dukaan's ChatGPT-powered customer service platform are already handling millions of interactions with high satisfaction rates. That's not a future scenario—that's now. The timeline for customer service roles accelerating toward AI replacement: 12-18 months for routine inquiries, 3-4 years for complex cases.
Routine coding and data entry face similar timelines. SHRM's 2026 Automation and AI survey shows that 32% of computer and mathematical roles were at least 50% automated in 2025, increasing to nearly 52% by 2026. That's acceleration in real time.
Management and strategic roles face far lower automation pressure because they require contextual judgment, stakeholder navigation, and decision-making that AI can inform but not replace. According to the WEF, roughly 41% of employers globally plan to reduce their workforce in areas where AI can automate tasks over the next five years. That means 59% are either growing in those areas or not planning reductions. The companies in that 59%? They're growing faster and paying more.
The Real Play: Become Antifragile in Your Current Role
You don't have to quit to position yourself for the 2025-2030 window. You can start right now where you are. Three concrete moves:
First, shift what you're accountable for. Stop measuring yourself on execution velocity. Start measuring yourself on outcomes that required judgment. In a field where your first tech job might already be gone, the jobs that remain are the ones where you influence what gets done, not just do it.
Second, build relationships inside AI-adopting companies. This doesn't mean interviewing tomorrow. It means coffee conversations with people working on AI infrastructure, AI-augmented workflows, AI implementation teams. These companies are hiring. They're paying more. They're growing faster. When you're ready to move in 12-24 months, those relationships turn into actual opportunities instead of cold applications.
Third, learn one skill your role will need in 2027. For developers, that's prompt engineering and working with AI agents. For sales teams, it's interpreting AI-generated leads and building relationships with people AI flagged as high-value. For operations, it's understanding what workflows AI can automate and where humans still add value. This isn't about becoming an AI expert. It's about becoming fluent in your field plus AI.
The antifragile move is positioning yourself so that AI adoption in your field makes your career better, not threatens it. That means being inside companies where AI is accelerating growth, understanding how AI changes your role, and building skills that let you work alongside AI instead of against it.
When AI Agents Become Fully Autonomous (And What That Means for Your Timeline)
True autonomy in the workplace isn't here yet. Even successful AI agent deployments like JPMorgan's Coach AI still operate within human oversight—accelerating research retrieval and increasing sales, but requiring wealth advisors to make final decisions. That pattern holds across industries: AI agents amplify what humans do instead of replacing the human entirely.
Full autonomy—where an AI agent makes decisions, executes work, and takes accountability without human review—is still 3-5 years away for most knowledge work. That's actually good news. It means the 2025-2030 window is a transition period, not a cliff. You have time to move, learn, and position yourself before full autonomy hits.
The companies that move fastest toward that autonomy are the ones where human workers shift into supervision, judgment, and strategy roles. That's where the 78 million emerging jobs live.
What I Tell People When They Ask If They Should Panic
The displacement is real. 77,999 tech jobs vanished in six months. Wall Street is planning to eliminate 200,000 jobs over the next 3-5 years. That's not alarmism—that's structural change accelerating.
But panic is optional. Here's what actually matters:
You have a two-year window before your current role either becomes antifragile or becomes at-risk. That means decisions made right now—about which company you're at, what skills you're building, who you're connected with—directly determine your career prospects in 2026-2028. If you're in a slow-adopting company, that's a flag. If you're building only execution skills, that's a flag. If you're not connected to people inside AI-forward organizations, that's a flag.
The honest ask: map your role onto automation risk. Know your number. If you're in customer service, operations, or routine coding, your timeline is compressed. If you're in management or strategy, you have more runway. Use that time to shift your output toward judgment, build relationships in growth companies, and learn how AI changes your field.
The 78 million emerging jobs exist. Companies deploying AI agents at scale are already hiring for the roles that emerge alongside them. The difference between moving into those roles and competing for jobs being shed comes down to timing and positioning. You're already inside the two-year window. The question is whether you're moving toward the growth or waiting for the squeeze.
