The AI Job Memo
The AI Career Playbook for Young Professionals
Last week, Dario Amodei said something that made every white-collar professional under 30 briefly consider a career in plumbing: half of all entry-level jobs in law, consulting, tech, and finance could be gone within five years.
The quote went viral. The takes wrote themselves. The ladder is disappearing. The bottom rungs are dissolving. Run.
I want to offer a different read.
As a AI Opportunity + PMF subscriber you’ll also get access to all the other Claude resources we put together, such as:
4️⃣ 🚨 The Ultimate AI Fundraising Copilot (Built With Claude)
1️⃣2️⃣ How I Rebuilt My SEO Strategy Inside Claude (The Playbook)
We’re already 70,000+ founders and operators inside!
and just for this week you can get AI + PMF at 30% Discount!
What’s actually happening
Amodei is right that the work junior analysts, associates, and consultants do today.
Drafting from templates, running comps, synthesizing 200 pages into 20 bullet points is exactly what language models are built to do. He’s not wrong about the capability curve. He’s not wrong that enterprise buyers are deploying Claude to do this work right now.
But “the work is being automated” and “the role is disappearing” are two different claims.
The work is being automated. The role is being rewritten.
The new job description
Here’s the version of the story that nobody went viral with: a junior lawyer who knows how to direct Claude through discovery, draft memos, cross-check citations, and pressure-test arguments doesn’t do less work than the 2019 version of that role. They do ten times more. They do it at a quality that used to require three years of experience. And they do it while the partner is still on the first call of the morning.
That person isn’t being replaced. They’re being promoted on day one, whether their title reflects it or not.
The juniors who disappear are the ones who treat AI like a threat to hide from, or a novelty to dabble with. The juniors who thrive are the ones who treat it like the most important tool they’ll ever learn to use — more important than Excel was for finance in the 90s, more important than Bloomberg terminals, more important than legal research databases.
What the bar looks like now
If you’re early in your career in one of these fields, the honest expectation has shifted. It’s not enough to be smart, willing, and good at following instructions. That was the old deal. The new deal is closer to:
Know your craft well enough to spot when the model is wrong. Know your tools well enough to get ten times the output without ten times the errors. Know your clients and colleagues well enough that your judgment — not your typing speed — is what people are paying for.
This is a higher bar than the one that existed five years ago. It’s also a more interesting job. The parts of junior work that were genuinely soul-crushing — the document review marathons, the deck-formatting death marches, the comps-building drudgery — those parts are going away. What’s left is the thinking.
What seniors should actually be doing
If you manage or mentor junior people, the Amodei quote isn’t a reason to stop hiring them. It’s a reason to hire differently and train harder.
The junior who can run an AI-augmented workflow at partner-review quality is worth more than three juniors who can’t. That’s a real shift in how teams should be structured, but it doesn’t mean teams shrink to zero. It means the ratio changes, and the training gets more deliberate.
The firms that quietly figure this out in the next eighteen months will pull away from the ones still arguing about whether AI belongs in the workflow at all.
The part worth sitting with
Amodei’s prediction reads as a warning because it sounds like mass unemployment. Read it again with fresh eyes and it’s something else: a forecast that the floor of what counts as competent professional work is about to rise dramatically.
That’s not a career-ending event. It’s a career-defining one, for people who take it seriously now rather than in 2029.
Five years is a long time. Long enough to get very good at something. Long enough to become the person on your team who everyone else quietly asks for help. Long enough to build the kind of leverage Amodei is describing — not to be displaced by it.
The question isn’t whether your job survives. It’s whether you do the work now to become the version of yourself that the new job requires.
Starting, as always, this week.
Cheers
Guillermo
OH PS: My friends from Newcomer just announced their annual AI Summit in London: Cerebral Valley. I went last year and it was amazing. Check it out!
