US Job Market Visualizer

Original Article ↗ Hacker News Discussion ↗

Methodology & Data Quality

  • Underlying employment and wage data comes from the US Bureau of Labor Statistics (BLS).
  • Several commenters say BLS data and projections lag reality and have a mixed track record (e.g., actuaries, pharmacists). Others note the collection process is systematic and better than most alternatives, but inherently backward-looking.
  • The AI exposure score is generated by a single LLM prompt. Many call this “vibes-based” and non‑rigorous, warning that it risks being mistaken for serious analysis.
  • Some worry policymakers or executives will overinterpret these scores despite explicit caveats on the site.

AI Exposure Ratings & Occupation Examples

  • The scoring often conflicts with on-the-ground experience: e.g., manual farm labor currently facing automation is rated very low exposure; childcare and teachers are rated high, which many see as unrealistic given the need for physical presence, discipline, and human contact.
  • The split between “Software Developers” (projected growth) and “Computer Programmers” (projected decline) triggers debate. Some trace this to old BLS distinctions (design vs implementation), others say the roles are synonyms in practice.

Labor Market Conditions & Immigration

  • Multiple commenters note that despite “much faster than average” BLS growth for software developers, many developers, especially juniors, struggle to find work.
  • Explanations discussed include economic cycles, influx of visa holders (H1B, L1, OPT), and definitional shifts in job categories.
  • There is a heated argument over whether high‑skilled immigration suppresses wages and job availability versus “lump of labour fallacy” counterarguments and the historical contribution of immigrants to growth.

Productivity, Surplus, and Inequality

  • A major thread: where automation/AI surplus goes.
  • One side argues most productivity gains from technology have gone to capital owners and the very rich; others say a broader upper‑middle class has also benefited.
  • Some model AI as a huge “total addressable market” over the wage bill; others note that if AI becomes commoditized, profits may be modest and broadly shared via cheaper services rather than concentrated.

Visualization Design & Accessibility

  • Several users like the idea of an interactive BLS browser but criticize execution:
    • Treemap hierarchy is shallow and makes specific roles hard to find.
    • Heavy reliance on hover makes it poor on mobile.
    • Red/green scales without luminance or pattern cues are nearly unreadable for colorblind users.

Broader AI Sentiment

  • Discussion polarizes between “AI is inevitable/transformative, surf or drown” rhetoric and pushback comparing this to earlier hype cycles (crypto, early internet) and “AI‑washing” of layoffs.
  • Even AI‑positive users emphasize that current LLMs are unreliable for unsupervised numerical analysis and should be treated as tools, not oracles.