download MP3 · single-anchor news read
download MP3 · two-host conversation
A Cognizant–Pearson study finds AI now performs 37% of entry-level tasks in India, above the 33% global average — the first rung of the career ladder is under measurable pressure. The broader pattern is the same one driving 2026's wave of AI-cited tech layoffs and Oracle's 21,000 cuts: tasks once handled by junior staff get absorbed into the model, and the remaining roles demand more judgment, faster.
AWS CEO Matt Garman told Platformer that Amazon is still hiring 11,000 interns and junior employees this year, calling the idea of replacing junior staff with AI "one of the dumbest things I've ever heard." Hiring criteria are shifting toward learning ability over fixed skill sets — early-career roles are not disappearing, but the bar is moving. Meanwhile, OpenAI's new Jalapeño inference chip opens future technical roles connecting software, data centers, and chip-level performance.
Cybersecurity training now has to cover the new enterprise attack surface that AI tools create: a fake skill that passed security scans and reached 26,000 users shows agent permissions, prompt injection defense, plugin vetting, and runtime monitoring are now baseline competencies. Information-as-attack-surface patterns — hidden content injection, cognitive state poisoning — should be in every agent-deployer's threat model.
Today's story is workforce bifurcation: companies still need junior talent, but only the kind that can work productively with AI, understand security risk, and adapt as infrastructure shifts. This week, pair domain knowledge with practical AI fluency, automation judgment, and at least one concrete agent-security skill — that combination is what survives the squeeze.