People call it the pyramid-to-diamond shift. AI is taking over a lot of routine entry-level work. Here’s what’s really happening, and how to come out ahead.
For decades, careers looked like a pyramid. Lots of entry-level people at the bottom doing routine work, fewer in the middle, a small group at the top making the big calls. You earned your way up by doing the bottom work first.
The diamond is the new shape a lot of strategists are describing. AI handles more of the routine, high-volume work that used to be everyone’s first job, so the bottom gets narrower. The wide part moves to the middle, where people do judgment work with AI doing the heavy lifting.
That means the old runway is shorter. You have to show judgment and real value sooner.
01
Young workers in AI-exposed jobs are falling behind
Stanford economists Erik Brynjolfsson, Bharat Chandar and Ruyu Chen track this using ADP payroll data. In their August 2026 update, employment for 22 to 25 year olds in the most AI-exposed jobs is 19% below where it would be if it had kept pace with less-exposed peers. (Stanford Digital Economy Lab)
02
But AI that helps people isn’t cutting jobs
Same study, other half of the story. In jobs where AI mostly complements workers instead of replacing tasks, employment is flat or rising, especially for experienced people. The authors call these early signals, not proof of cause.
03
Hiring is slow for everyone
LinkedIn says overall hiring is down about 20% since 2022, and it mostly blames interest rates, not AI, for now (TechCrunch). Its 2026 Grad’s Guide has entry-level hiring down 6% year over year and mid-level down 10% (LinkedIn).
So here’s my read. The bottom rung is shrinking in jobs AI can do on its own. The people pulling ahead use AI to do more, not to be replaced by it. That’s the move.
01
Treat junior work as your leverage
First drafts, research, data cleanup and routine emails used to take a team. You can now do them solo with Claude. Show up already able to produce what used to take 3 people.
02
Build a portfolio of outputs
Don’t wait for a manager to vouch for you. Show the analyses, systems and client materials you actually made. AI helps you build real proof fast, even early in your career.
03
Learn to direct AI
The new junior skill is direction. Brief AI well, judge its output hard, and push it until the work is actually good. Most people never practice this. You should.
04
Get to judgment work faster
The old path was years of routine work before anyone gave you the interesting stuff. Let AI handle the routine part, then use the time you get back to ask harder questions, take on work above your level and spend time with senior people.
05
Make your AI skills visible
Name the tools. Show the work. Say it with numbers, like “I built a client research system in Claude that cut our prep time from 3 hours to 45 minutes.” Specific sentences like that get people hired.
01
Hand off now
02
Develop fast
03
Protect and grow
Paste this into Claude or ChatGPT, fill in the brackets and attach your resume if you have one.
01
Contextual judgment
Reading what’s really going on in a room or a client relationship. AI only knows what you tell it. Build it: get into the rooms where decisions get made.
02
Earned trust
The person who delivers every time. That track record compounds. Build it: under-promise, over-deliver.
03
Taste
Knowing what good looks like and lifting AI output to it. Build it: study the best work in your field and form strong opinions.
04
Hard conversations
Handling a difficult client or giving feedback that lands. AI can draft the words, but it can’t read the room. Build it: stop avoiding the hard talks.
05
Connecting the dots
Pulling ideas from different teams and fields into something new. Build it: read widely and work across functions.
06
Ownership
Being the person whose name is on the outcome. Build it: say “I’ll take that,” then follow through.