Planet Money – “Asking for a Friend… Which Jobs are Safe from AI?”

Release Date: September 10, 2025
Hosts: Sally Helm & Amanda Aronczyk (NPR)

Episode Overview

This episode tackles the growing anxiety about artificial intelligence (AI) and its impact on the future of work. Listeners and the Planet Money team ask the pressing question: Which jobs are actually safe from AI?
Through personal stories and expert interviews, the hosts examine new research frameworks that reveal not simply which positions may be “safe,” but how society’s entire relationship with work could change. The conversation moves from individual concerns to sweeping economic implications, ultimately suggesting that our response and imagination might matter as much as any list.

Personal Anxieties: Real Stories from Listeners

Time: 00:24–04:59

• Sally and Amanda introduce Charlie Baker, a college senior who falls in love with the LSAT but hesitates to pursue law out of fear that AI might automate legal work.
  • “I don’t know if it’s worth the investment now to go to law school for three years if I’m potentially going to just be replaced by an AI chatbot.” – Charlie Baker (01:57)
• Anna Wynn, a tech product designer, is also anxious as automation threatens her field.
  • “What would AI not automate out? …anything right now, physical, maybe a plumber…” – Anna Wynn (02:20, 03:01)
  • She considers switching to hands-on work, like welding or her mother’s job as a nail tech.
• The hosts highlight a wave of similar worries from listeners—people considering yoga teaching, real estate, or legacy jobs for safety.
  • “It feels to me like we all have no idea how to think about this…” – Sally Helm (04:42)

The Search for an Answer: Job “Exposure” to AI

Daniel Rock’s “Exposure” Framework

Time: 07:00–17:44

• Sally seeks a definitive list of jobs immune to AI and finds Daniel Rock’s research, done with OpenAI, which ranks 1,000+ occupations by “AI exposure.”
• How the list works:
  • Every job is broken into tasks (from the ONET database – 08:24).
  • Each task is scored for its exposure:
    • E0: Not exposed (AI can’t help)
    • E1: Fully exposed (AI can do the task quickly)
    • E2: Somewhat exposed (AI might help with added systems)
  • “Exposure to AI is not the same thing as this job will be automated.” – Sally Helm (12:17)
• Examples:
  • Acute care nurses: “administer blood transfusions” = E0; “document data” = E1
  • High-exposure (more likely to change): Public relations, translators, knowledge workers
    • “She had it [GPT-4] write a press release for her in her tone and she said it did an absolutely great job… this is the first few years of my career just in a machine.” – Daniel Rock (11:20)
  • Low-exposure (less likely to change): Welders, short order cooks, athletes
• Rock’s Argument:
  • Exposure means potential for change—not a hit list.
  • AI is like electricity: a “general purpose technology” that will shift work in unexpected ways (13:11).
    • “So much is going to change that we really can't say which jobs are going away, which jobs are going to become more important.” – Daniel Rock (14:24)
  • Some fields may grow because of AI, not in spite of it.
    • “If you're really exposed, it could be great for you if you could use AI to make yourself a thousand times more productive… you might do really well.” – Daniel Rock (15:31)
• Key takeaway: The most-exposed jobs will see the most change, but not necessarily elimination. Demand elasticity, adoption patterns, and social choices all matter.
  • “We just can’t know.” – Daniel Rock (28:51)

The “Humanness” Factor: The EP.O.C.H. Framework

Isabella Loisa’s “Epoch Score”

Time: 19:58–27:59

• Researcher Isabella Loisa (MIT) takes a different approach: Instead of AI’s exposure, she asks what skills humans bring that AI can’t (yet) match.
  • She responds to widespread fear:
    “I don’t think I was ever afraid of AI taking over my job. I was more perhaps afraid of the fear that people were feeling.” – Isabella Loisa (20:08)
• The “Ep.O.C.H.” Score:
  • E: Empathy (genuine human connection)
  • P: Physical Presence (need to be there in person)
  • O: Opinion/Judgement/Ethics (critical thinking, moral calls)
  • C: Creativity (inventing new things)
  • H: Hope/Vision/Leadership (inspiring, guiding, motivating)
• Each occupation is scored; those with higher Epoch values are more “human-shaped.”
• Examples:
  • Emergency management directors and managers score high (empathy, presence, judgment).
  • Construction workers score high—surprisingly, because of mentoring others.
    • “Turns out… they are mentoring or teaching less experienced construction workers.” – Isabella Loisa (24:31)
  • Clerical jobs (tax preparers, insurance appraisers) score low.
• The complexity of real jobs: Many mix “automatable” and “human” tasks. So even within one field, some tasks could be replaced by AI, some merely empowered.
  • “If you have a bunch of clerical tasks, but also some connected tasks that are highly human, then your job might be safer.” – Amanda Aronczyk (26:08)
  • For lawyers: writing briefs (AI-possible), courtroom argument (human-essential).
    • “All the different occupations that require critical thinking, judgment, even creativity, that is not going to go away.” – Isabella Loisa (27:19)
• Advice:
  • Focus on mastering the “human” skills.
  • “Like learn how to think.” – Isabella Loisa (27:56)

Navigating the Uncertain Future

Concrete Advice & Real-Life Example

Time: 28:58–31:50

• There’s no one “safe” career, and there may never be a neat master list.
• Both Rock and Loisa agree:
  • “Learn to use AI so that you can be ready… shape it to your advantage.” (28:58)
  • But: This is hard, requires trial and error, and ethical choices.
• Example: Veterinarian Kat Reardon
  • Used AI (ChatGPT) to draft patient notes, saving time and reducing risk of injury (since both hands stay on the animal).
    • “I’ll get bit less often … I can kind of feel if they’re kind of starting to get upset … because I was distracted and not paying attention previously.” – Kat Reardon (30:44)
  • She transitions part-time into teaching AI to help other vets.
  • “It’s the AI augmentation story… and not one that I would have imagined.” – Sally Helm (31:09)
• Big Picture:
  Imagination and adaptability may be the most important skills of all. Change is likely to be surprising—and we’re writing the rules as we go.
  • “I went in looking for a list… but I learned there really is no list. Not yet. Maybe not ever. We are in for a weirder ride than that.” – Sally Helm (31:09)

Notable Quotes and Moments

| Time | Speaker | Quote | |---------|---------------------|-----------------------------------------------------------------------------------------------------------| | 01:57 | Charlie Baker | “I don’t know if it’s worth the investment now to go to law school… if I’m potentially… replaced by AI.” | | 04:42 | Sally Helm | “It feels to me like we all have no idea how to think about this…” | | 11:20 | Daniel Rock | “She had it [GPT-4] write a press release… She said, wow, this is the first few years of my career in a machine.” | | 12:17 | Sally Helm | “Exposure to AI is not the same thing as this job will be automated.” | | 14:24 | Daniel Rock | “So much is going to change that we really can't say which jobs are going away…” | | 15:31 | Daniel Rock | “If you’re really exposed, it could be great… you might do really well.” | | 20:08 | Isabella Loisa | “I was more perhaps afraid of the fear that people were feeling.” | | 27:19 | Isabella Loisa | “All the different occupations that require critical thinking, judgment, even creativity… not going away.”| | 27:56 | Isabella Loisa | “Like learn how to think.” | | 28:51 | Daniel Rock | “We just can’t know.” | | 30:44 | Kat Reardon | “I’ll get bit less often … because I was… not paying attention previously because I was getting my notes done.”| | 31:09 | Sally Helm | “I went in looking for a list… but there really is no list. Not yet. Maybe not ever.” |

Key Insights and Takeaways

• No job is absolutely immune—AI is a “general purpose technology” that will shape, eliminate, and create jobs in unpredictable ways.
• Exposure ≠ Elimination: Being exposed to AI mainly predicts how much your job is likely to change, not whether it will vanish.
• “Humanness” matters: Roles heavy on empathy, ethical judgment, creativity, physical presence, and hope/leadership are less likely to be fully automated.
• Augmentation, not just automation: Many jobs will be transformed, not replaced, as AI takes over certain tasks and enhances others.
• Adaptability is crucial: Those who actively learn to use AI can shape its adoption in their fields.
• Imagination and flexibility are as important as technical skills in the looming age of AI.

Further Resources

• Daniel Rock’s research: AI “exposure” across the labor market
• Isabella Loisa’s research: The “Ep.O.C.H.” humanness framework

For those anxious about AI’s impact on work:
There’s no “magic list” of safe jobs—but the more human your work, the more personally adaptable you are, and the more you proactively engage with new technology, the more likely you are to thrive. The future of work is not fixed; it’s ours, collectively, to invent.