The "AI kills software jobs" framing and the "AI is just a tool" framing are both wrong in the same way — they treat it as uniform.
What actually seems to be happening is that AI compresses the value of generic output and amplifies the value of domain judgment. A developer who knows how a specific industry works, what the edge cases are, why the legacy system was built the way it was — that context isn't in the training data. It compounds.
The engineers I've seen struggle most aren't the ones in AI-exposed roles. They're the ones in AI-exposed roles who've optimized for output volume rather than judgment depth. Those two things used to correlate. They no longer do.
The feedback loop described here is what stuck with me — AI improves, companies cut headcount, savings go back into AI, AI improves. No natural brake.
The article puts a specific number on it: a $180K PM replaced by a $200/mo AI agent. I've been building a tool that lets you run this kind of scenario on your own career — scores your AI exposure and simulates paths that reduce it.
One thing I've found from running hundreds of simulations: augmenting your current career with AI consistently leads to better financial outcomes over 5-10 years than pivoting to a new field entirely.
The best move isn't to run — it's to adapt in place.
Free to try: parallaxapp.world
My son had questions about his career and couldn't answer them. So I built Parallax — a tool that helps you explore career options and arrive at your own answers.
Upload your resume, get an AI exposure score (0-100), then run simulations to see what happens if you: stay and upskill, pivot to adjacent work, or reset entirely. Year-by-year timelines. Trade-offs made visceral.
* The problem
People know AI is disrupting careers. What they don't know is:
How exposed they specifically are, given their actual career history
What their realistic options are — not generic advice, but paths modeled on their timeline
What each option costs them in money, stress, and life satisfaction
* What Parallax does
It's a flight simulator for your career. Not advice — consequence exploration.
The flow (takes ~5 minutes):
* Upload resume → AI calculates your AI exposure score
* Five-question interview (finances, learning appetite, stability preference)
* LLM generates up to 9 futures: 3 strategies × 3 timing scenarios
* Pick which paths to compare
* Compare dashboard shows net worth, stress, fulfillment, AI exposure, drift
* Three recovery strategies:
1) Augment (blue): Stay, upskill with AI tools
2) Pivot (amber): Lateral move, lower AI exposure
3) Reset (purple): Start something fundamentally different
* Tech
Next.js 14 + TypeScript, LLM-powered
No database — fully client-side (localStorage)
Zero tracking. Your career data never leaves your browser.
Privacy-first by design
* Pricing
Free: AI exposure score + interview + briefing
Pro ($10/mo): Full compare dashboard + timeline exploration
This app is compute intensive, so I added a paywall to pay for my compute costs.
Built solo over 2 months. Would love your feedback.
What actually seems to be happening is that AI compresses the value of generic output and amplifies the value of domain judgment. A developer who knows how a specific industry works, what the edge cases are, why the legacy system was built the way it was — that context isn't in the training data. It compounds.
The engineers I've seen struggle most aren't the ones in AI-exposed roles. They're the ones in AI-exposed roles who've optimized for output volume rather than judgment depth. Those two things used to correlate. They no longer do.
reply