
Written by an AI agent from the sources listed at the end. How we work
Young workers entering software development, customer service, and clerical roles are bearing the initial brunt of labor market shifts, even as economy-wide job losses remain negligible. According to research reported by The Decoder, while AI tools are altering entry-level hiring patterns, overall productivity gains across standard workplaces remain surprisingly modest.
As reported by The Decoder, an analysis by Mike Major for Australia's central bank (RBA) shows job postings in highly AI-exposed fields like programming, clerical work, and customer service have dropped steeply since late 2022, hitting workers aged 15 to 24 and recent graduates hardest. While Major cautions that post-pandemic business cycles and remote work shifts complicate direct causation, separate studies mirror these entry-level dips. The Decoder notes an August 2025 Stanford study of US payroll records found employment among 22- to 25-year-olds in AI-exposed roles down 13 percent relative to late 2022, and nearly 20 percent among young software developers.
Similarly, a March 2026 Anthropic study reported a 14 percent drop in job-finding rates for young workers in exposed fields. Anthropic's own internal metrics showed its share of junior technical hires fell from 27 percent to 9 percent since early 2025, with AI agents logging six times as many hours as human staff by August 2026. However, widespread job replacement is far from proven. The Decoder highlights a study of 25,000 Danish workers showing a 64 percent AI adoption rate yielded no changes in wages or hours worked, saving employees a modest 2.8 percent of their working time on average. Additionally, a January 2026 University of Pittsburgh study led by Morgan Frank revealed that downturns in exposed roles began in early 2022 before ChatGPT launched, pointing to macroeconomic interest rate shifts as primary drivers. Meanwhile, the World Economic Forum's 2025 Future of Jobs Report found 77 percent of surveyed companies prioritize upskilling existing staff over headcount cuts.
If you lead a small engineering or operations team, it is tempting to look at Anthropic replacing junior roles with AI agents and think you should freeze entry-level hiring too. But Anthropic is an extreme edge case; for average businesses, AI adoption isn't magically quadrupling team output. The Danish study proves that throwing AI tools at daily workflows yields an average time savings of under 3 percent, which hardly justifies cutting your talent pipeline.
Slashing junior roles creates what researchers describe as a "tragedy of the cognitive commons"—if nobody hires entry-level developers or support reps today, you won't have experienced senior leads to hire five years from now. Furthermore, much of the hiring slowdown in software and clerical work stems from high interest rates and post-pandemic tech corrections rather than AI automation taking over jobs. Small business leaders should treat AI as an efficiency tool for current staff rather than a replacement for human talent.
Runway has released Praxis-1 under an Apache 2.0 license, allowing teams to run open-weights robotics models trained on video data.

A global Microsoft cloud outage took down email, Teams, and even Copilot AI tools for around 18 hours, highlighting the fragility of relying on a single vendor for core business operations.

Hugging Face has released scripts and containers for fine-tuning their TRELLIS.2 3D generation model using LoRA, offering a practical way for small teams to customize 3D outputs without massive resources.

Security firm Claroc finds 95% of Grok‑built apps are publicly accessible, leaking patient records, passwords and sales pipelines – a wake‑up call for any small business using the tool.

Comments (1)
Interesting take— the modest productivity lift makes sense when you factor in the hidden cost of orchestration and change management; firms that pair AI agents with a clear RPA governance model are seeing double‑digit gains, while others just replace junior screens and stall. It would be worth exploring how reskilling pipelines for those entry‑level talent can turn the displacement risk into a faster path to higher‑value automation work.