Divya van Mahajan

Is AI Already Shrinking Entry-Level Tech Jobs?

ai-daily-brief-podcast

Overview

This episode of the AI Daily Brief (dated 2025-05-29) examines whether AI is already measurably reducing entry-level tech hiring, drawing primarily on a 2025 State of Talent Report from venture capital firm SignalFire. The episode also covers three headline news items: Anthropic’s launch of voice mode for Claude, Meta’s restructuring of its AI division, and OpenAI’s exploration of a “Sign in with ChatGPT” feature. The host (name not stated) contextualises the hiring data within broader economic trends and offers a cautiously optimistic but clear-eyed assessment of the transition ahead.

Source video: URL not provided.

Prerequisites

  • Basic familiarity with the AI industry landscape and major players (OpenAI, Anthropic, Meta, Google/DeepMind, Amazon)
  • General understanding of tech hiring cycles and the post-pandemic boom-and-bust in software engineering jobs
  • Awareness of concepts such as large language models (LLMs), agentic AI assistants, and foundation models
  • Familiarity with the term “Mag7” (the seven largest US technology companies by market capitalisation)

Main Points

Headline: Anthropic Launches Voice Mode for Claude

  • Claude’s voice mode allows spoken input and audio responses, with five voice options; the default has a British accent
  • Anthropic is positioning the feature as a step toward a full agentic assistant (e.g., checking a calendar, drafting emails)
  • Free users are limited to approximately 20–30 voice conversations; tool-use integrations (calendar, email) require a paid subscription
  • At launch, voice mode is only available via the Claude app, not the web interface or API
  • The host frames this as “table stakes” given ChatGPT’s existing voice capabilities, but notes interest in Anthropic’s agentic framing

Headline: Meta Splits Its AI Division in Two

  • An internal memo from Chief Product Officer Chris Cox announced the Gen AI division will be divided into an AI Products team (led by Connor Hayes, covering Meta AI, AI Studio, and in-app tools) and an AGI Foundations unit (co-led by Ahmad El-Dale and Amir Frankel, focused on foundation Llama models)
  • The Fundamental AI Research (FAIR) lab remains a separate division, though one multimedia team moves to AGI Foundations
  • No executives are leaving and no jobs are being cut; the goal is faster product development and greater organisational agility
  • This contrasts with Google’s strategy of consolidating AI teams under DeepMind to bring research closer to product
  • Context: Meta had previously reorganised in 2023 (moving Llama out of FAIR), and recent reporting described a “brain drain” to companies like Mistral

Headline: OpenAI Explores “Sign in with ChatGPT”

  • OpenAI is soliciting developer interest in a feature allowing users to authenticate into third-party apps using their ChatGPT account
  • The initiative targets apps ranging from 1,000 users to over 100 million; OpenAI is also asking whether apps use OpenAI APIs and how they monetise AI features
  • A preview launched earlier in May via Codex CLI, with free API credits as an incentive
  • Observers compare it to Facebook’s early platform strategy: users could bring API tokens, GPTs, memories, and tools into third-party ecosystems
  • CEO Sam Altman has discussed “Sign in with OpenAI” since at least late 2023; his WorldCoin project reflects a parallel interest in unified biometric credentialing

Main Topic: Is AI Already Shrinking Entry-Level Tech Jobs?

SignalFire’s State of Talent Report 2025

  • SignalFire’s Beacon platform tracks 650 million professionals across 80 million organisations
  • Key finding: entry-level hiring is “collapsing,” and the decline appears to go beyond post-pandemic normalisation
  • Between 2023 and 2024, every demographic with two or more years of experience saw hiring increases at both startups and big tech; entry-level hiring fell 25% at big tech in 2024
  • New grad hiring at big tech is down over 50% from pre-pandemic (2019) levels; for broader new hires, down over 30% from 2019
  • The total share of hires that are entry-level has approximately halved across both startup and big tech categories

Broader Labour Market Signals

  • Federal Reserve Bank of New York data: recent college graduate unemployment is 5.8%, vs. 4% for the general population — the highest since 2013 (excluding the pandemic) and the first sustained multi-year upward trend since the data series began in 1990
  • Law School Admissions Council: 2025 law school applications up ~21% year-over-year, consistent with recession-era patterns of deferring workforce entry
  • Only ~70% of computer science graduates from the class of 2024 found employment within six months; only 61% of those were employed as engineers, and just 12% landed at Mag7 companies — all figures at or near five-year lows

Causes: Macro vs. AI

  • SignalFire acknowledges multiple drivers: the end of low-interest-rate “free money,” overhiring during 2020–2022, tighter funding (Carta data shows Series A startups are 20% smaller than in 2020), and shorter runways
  • However, SignalFire also attributes part of the shift to AI automating routine, entry-level tasks; big tech is prioritising machine learning and data engineering while non-technical functions (recruiting, product, sales) shrink
  • World Economic Forum Future of Jobs report: 40% of employers plan to reduce headcount in areas where AI can automate tasks; entry-level roles in finance and consulting are also declining

The “Doing More” Problem

  • The host draws on a post by Aaron Francis noting that AI assistance has increased, not decreased, the number of coding tasks he faces (“we added three lanes to the highway and still have traffic”)
  • A New York Times report on Amazon describes engineers whose teams have halved in size but are expected to produce the same volume of code using AI; managers are raising output goals and tightening deadlines
  • A Harvard labour economist characterises this as a “speed-up for knowledge workers,” raising the question of how productivity gains from AI are distributed between employer and employee

Leadership, Culture, and the Early Career Mentorship Gap

  • The host argues that where AI’s impact falls on a spectrum from “dehumanising to rehumanising” depends largely on leadership decisions and how those decisions are communicated to employees
  • If entry-level hiring does not recover, an entire generation may lack the mentorship and on-ramp traditionally provided by junior roles
  • The World Economic Forum suggests AI could instead be used to democratise access and accelerate training of junior workers into senior roles, potentially through apprenticeship models
  • The host’s personal takeaway: identify and hire talent that larger companies are overlooking, while acknowledging the transition will be disruptive regardless of long-term optimism

Key Concepts

  • SignalFire Beacon: A data platform tracking over 650 million professionals and 80 million organisations, used to analyse hiring trends
  • Entry-level hiring collapse: The observed sharp decline in hiring of recent graduates and workers with fewer than two years of experience, particularly in tech
  • Agentic voice assistant: An AI voice interface capable of taking multi-step actions on behalf of a user (e.g., reading a calendar, sending emails), beyond simple question-and-answer
  • Sign in with ChatGPT: A proposed OpenAI authentication feature allowing users to log into third-party apps via their ChatGPT account, analogous to “Sign in with Google/Facebook”
  • AGI Foundations unit: Meta’s newly created division focused on long-horizon research and improving foundation Llama models, distinct from its product-facing AI team
  • FAIR (Fundamental AI Research) lab: Meta’s long-standing foundational AI research division, remaining separate from the newly restructured product and AGI teams
  • Speed-up dynamic: The phenomenon whereby AI productivity tools lead employers to increase output expectations proportionally, rather than reducing workload or headcount costs for employees
  • Early career mentorship gap: The risk that a sustained decline in entry-level hiring eliminates the traditional on-ramp through which junior workers develop skills under senior mentorship

Summary

The episode’s central argument is that while AI’s impact on employment has largely been theoretical, 2024 and 2025 data are beginning to reveal a concrete and measurable effect: entry-level and new graduate hiring in tech — and increasingly across other professional sectors — is declining sharply, and the decline cannot be fully explained by post-pandemic macroeconomic normalisation alone. Drawing on SignalFire’s 2025 State of Talent Report, Federal Reserve data, and World Economic Forum findings, the host presents evidence that AI is automating the routine tasks historically assigned to junior workers, causing companies to hire fewer entry-level employees while increasing expectations for experienced staff. The host is net optimistic about long-term outcomes — anticipating that demand for software and AI-enabled work will expand to absorb displaced talent in new forms — but stresses that the transition will be painful and uneven, particularly for recent graduates who may miss the formative mentorship that entry-level roles provide. The episode closes with a call for deliberate leadership: companies and managers must actively decide how to distribute AI productivity gains and whether to invest in developing the next generation of workers, rather than simply compressing more output from a smaller, more experienced workforce.