TED Radio Hour (NPR) — Episode Summary

Title: How companies use AI to choose who gets hired and fired
Air Date: October 3, 2025
Host: Manoush Zomorodi
Guest: Hilke Schellmann, investigative journalist, author of The Algorithm and professor at NYU

Overview

This episode investigates the profound influence of artificial intelligence (AI) on hiring, monitoring, and firing in the modern workplace. Host Manoush Zomorodi and guest Hilke Schellmann explore how AI’s growing adoption in recruitment and people analytics is reshaping job markets—promising fairness and efficiency, yet often increasing opacity, anxiety, and bias for employees. Interwoven are real-world stories, technical analysis, and practical advice for navigating this new world of work.

Key Discussion Points & Insights

1. The Evolution of Hiring — From Classifieds to AI (01:24–04:47)

• Hiring used to be personal and paper-based: "Back about 25 years or so ago...people still looked through the classifieds. They mailed or emailed in their resume." (A, 01:24)
• Job platforms (LinkedIn, Monster, ZipRecruiter) democratized access but created application volume overload, making hiring overwhelming for employers. (A&C, 02:05–02:58)
• Software, then AI, stepped in to sift through resumes.
  • Example: Amazon’s AI analyzed resumes to predict future job success but amplified hidden gender biases.
  • "The algorithm assumed that men would be best suited for the job... started to penalize people who had the word 'woman' or 'women' on their resume." (C, 04:47)
  • Amazon shelved the tool after attempts to fix it failed.

2. The Dangers and Opacity of AI in Hiring (05:28–11:24)

• Opacity and legal risks lead companies to hide AI failures. (C, 05:28–06:25)
  • Fear of massive liability keeps AI hiring processes secretive.
  • "It's really hard for applicants, right? ... it's sort of like a little bit of a cloak of silence here.” (C, 06:18)
• Move toward AI tools for one-way video interviews, facial analysis, and more.
  • Hilke describes attending a demo in 2018 where AI claimed to analyze facial expressions, voice, and language for job suitability.
  • Expert psychologists debunk the science, calling it “pure rubbish” and warning of increased bias. (C, 08:09)

3. Algorithms at Work: Beyond Hiring (09:54–13:34)

• AI now permeates the full ‘employee lifecycle’:
  • Used in background checks, video interviews, and ongoing workplace surveillance.
  • Monitors keystrokes, Zoom meetings, sentiment in emails—potentially fueling firings based on non-skills-related factors.

4. Real-world Consequences: Lizzie's Story (16:37–19:06)

• Lizzie, a UK makeup artist, was laid off after a pandemic-mandated one-way video interview scored by AI.
  • She received no feedback, was told only she scored too low, and felt “shattered.”
  • She and two colleagues eventually settled a lawsuit with the company. (C, 16:37–19:06)
• Hilke’s own tests of these interview tools showed alarming unreliability:
  • Responded to all questions with “I love teamwork” and still earned good scores. (C, 19:54–20:35)
  • Answered in German for an English job and scored 73%.
  • "I was just surprised when I got a result... this makes no sense... I just said, you know, basically just make sounds for the tool." (C, 21:05)

5. AI-Powered Monitoring and Productivity "Theater" (22:42–29:54)

• Surveillance increasingly normalized across industries.
  • At least 8 of the 10 largest US companies monitor employees—tracking keystrokes, webcam activity, time on tasks, and more. (A&C, 23:02–23:17)
  • Case: Medical coder Emily Smith monitored intensely, developed anxiety, set timers for breaks to avoid red-flags.
  • Rise of 'Productivity Theater': Employees send early/late messages, use “mouse jigglers,” and join irrelevant meetings to appear active.
    • “Microsoft found out that...people engage about an hour a day on productivity theater.” (C, 26:51)
• Productivity metrics rarely capture true value, creativity, or critical thinking.
  • "You can only often only track signals that happen on your computer." (C, 29:54)
  • Soft skills, collaboration, and out-of-box thinking are often ignored by AI’s blunt measures.

6. AI in Firing & Layoffs — Flight Risks and Bias (36:46–40:07)

• AI predicts “flight risk” and productivity for retention and layoffs.
  • Can flag an employee for leaving based on printing history, lack of promotion, etc.
  • Firings in warehouses and gig work sometimes fully automated based on algorithmic productivity scores.
• Biased data feeds into People Analytics:
  • Women and people of color rated less successful in performance reviews, bias propagates through AI tools.
  • “It looks like a perfect score... but the bias now gets objectified in these tools and gets buried in the math of it.” (C, 38:50)

7. The New Reality: Human Skills & Automation (40:07–45:14)

• Companies still crave “human” skills: creativity, judgment, care—but struggle to assess them.
• The future of work: AI will take over some tasks but not whole jobs; synergy will outpace dystopia.
  • “We are moving away from this dystopian idea that AI will take over all our jobs... right now our workload has actually increased...” (C, 42:24)

8. Navigating AI as an Applicant: Practical Tips (47:50–52:56)

• Tip 1: Keep your resume “boring” but machine-readable. Avoid fancy formatting, images, multiple columns, or special characters. Use standard templates. (A&C, 48:10)
• Tip 2: Use clear, quantifiable achievements. E.g., “Saved company $5 million.”
• Tip 3: Include keywords from the job description, but don’t copy it verbatim. 80–90% keyword overlap is a good target.
• Tip 4: List all your skills—hard and soft. Use bullet points, create a dedicated skills section.
• Tip 5: Don’t hesitate to use AI for resume polish (grammar, clarity).
• Tip 6: Apply directly on the company website for better chances—internal applicant pools are prioritized.
• Tip 7: Leverage human connections. Employee referrals or engagement with recruiters on job platforms bump you past early AI screening.
  • “If an employee currently at a company recommends you... you often bypass... the first phases of rejection by AI tools.” (C, 51:20)

Notable Quotes & Memorable Moments

• On the illusion of AI fairness:
  “The algorithm assumed that men would be best suited for the job... started to penalize people who had the word 'woman' or 'women' on their resume.” — Hilke Schellmann (04:47)

• On black-box hiring:
  "It's sort of like a little bit of a cloak of silence here." — Hilke Schellmann (06:18)

• On misleading AI assessments:
  "I literally just said, you know, basically just make sounds for the tool, right? ... I was 73% qualified for the job. And I was like, what the—this makes no sense." — Hilke Schellmann (21:05)

• On productivity tracking:
  “Microsoft found out that...people engage about an hour a day on productivity theater.” — Hilke Schellmann (26:51)

• On bias in automated reviews:
  “All of this data... is biased. So then the bias now gets objectified... and gets buried in the math of it.” — Hilke Schellmann (38:50)

• On adapting resumes for AI:
  “Use short, crisp sentences, be declarative, and quantify achievements... you didn’t save money, you saved the company $5 million.” — Hilke Schellmann (48:58)

• On networking in the AI age:
  "You often bypass sort of the first phases of rejection by AI tools and you, like, land directly on the desktop for a recruiter or hiring manager.” — Hilke Schellmann (51:20)

Timestamps for Important Segments

• 01:24 — The shift from classifieds to online platforms
• 04:47 — Amazon’s failed AI hiring tool and gender bias
• 06:18 — Secrecy and legal fears around AI hiring failures
• 08:09 — Debunking the science behind facial analysis in hiring
• 16:37 — Lizzie’s story: Losing her job due to a low AI video score
• 19:54 — Testing and fooling one-way interview algorithms
• 23:02 — Pervasive monitoring: keystrokes, webcam, activity
• 26:51 — “Productivity theater” and mouse jigglers
• 36:46 — AI-driven layoff criteria and performance reviews
• 38:50 — The problem of bias in People Analytics
• 47:50 — Applicant survival tips in AI-driven job markets
• 51:20 — Importance of networking and recruiter engagement

Episode Tone & Takeaway

The tone is cautionary but pragmatic, blending investigative skepticism with real-world hope. Schellmann and Zomorodi urge listeners to stay curious, ask questions about AI’s role in their work lives, and adapt practically—whether as job seekers, workers, or leaders. AI is not going away but can, if scrutinized and improved, become a tool for good rather than harm.

For further learning:

• Watch Hilke Schellmann’s TED Talk at TED.com
• Consult The Algorithm: How AI Decides Who Gets Hired, Monitored and Fired — and Why We Need to Fight Back Now for a deeper dive.