Anthropic CEO Dario Amodei caused commotion, at least in my corner of the internet, when he told Axios that “AI may wipe out 50 percent of entry-level white-collar roles within five years.”
Fresh data suggests he might be right:
- Big Tech graduate hiring is down by 50% since 2019 (SignalFire).
- Meta’s Mark Zuckerberg said AI will soon handle “mid-level engineer” work.
- ServiceNow agents already resolve 80% of support tickets unaided.
- An EY survey reveals that 48% of tech executives are deploying agent-based AI.
- Salesforce is redeploying 500 service reps into “data-plus-AI roles.”
Together, the signals point to a “broken bottom rung on the career ladder,” as LinkedIn’s Chief Economic Opportunity Officer, Aneesh Raman, wrote in The New York Times:
“In tech, advanced coding tools are creeping into the tasks of writing simple code and debugging — the ways junior developers gain experience. In law firms, junior paralegals and first-year associates who once cut their teeth on document review are handing weeks of work over to A.I. tools to complete in a matter of hours. And across retailers, A.I. chatbots and automated customer service tools are taking on duties once assigned to young associates.” – Aneesh Raman, Chief Economic Opportunity Officer, LinkedIn
So what does the entry-level job of the future look like, and how do we (workers and companies) get there?
Let’s dive in.
The New Entry-Level Job
I liked Josh Bersin’s counter to Amodei’s prediction that entry-level jobs are about learning and building a pipeline, not just cheap labor. And that cutting them creates long-term skill gaps.
This is something LinkedIn’s Aneesh Raman also highlighted: getting started later means less money and more inequality.
I’m also puzzled by what entry-level jobs are, anyway. Any first job that people (eventually) take is entry-level. How can they go away?
But yes, if we interpret those as the kind of roles Raman describes, I do think Amodei’s “50%” is directionally right. Jobs where you’re just copy-pasting, summarizing, or creating first drafts will become obsolete with AI.
However, in its place will be ‘first roles’ as ‘AI leaders' that young candidates are ideally suited for.
They enter the workforce as AI natives, and in fact, can teach today's leaders a lesson or two (reverse mentoring).
This is a place where graduates can actually have a significant advantage: Sam Altman recently remarked that younger AI users often tap it as a new kind of operating system, rather than a Google replacement.
These are the kind of people that employers are hungry for, as evidenced by Zapier CEO Wade Foster, who stated that 100% of new hires must be fluent in AI.

Because, as I recently stressed in my articles Prompt Engineering Isn’t Dead and You’re Only Using 10% of ChatGPT, AI still needs direction.
Today’s models are incredibly powerful. I’ve been working more with o3, and it amazes me what a reasoning model + tools can do. But no matter how impressive, they still need direction.
“The machine can do anything, now what do you want them to do?”
Tapping AI as a team member and prompting them well is key, alongside proficiency in tool fluency, to start or continue any career.
Even Amodei told Anderson Cooper so: the best hedge to future disruptions is simply “learn AI.”
That message is loudest for fresh graduates who now know what employers value:
- Baseline expectation: Do you know how to prompt, set up a simple assistant, and leverage different models? (See my new ChatGPT course if you need help.)
- Become a SuperWorker: Use my GED-RT framework (General, Error-friendly, Digital, Recurring, Toil) to identify tasks that AI (beyond ChatGPT) can handle today.
- Show real projects: custom GPTs, automations, and prompt libraries: proof you can tailor these models for real productivity and impact. (the stuff we do in the Lead with AI executive bootcamp.)
Do so, and graduates may even agree with PwC, which said they’re “lucky to be starting their careers as AI accelerates.”