AbovetheNoise

Home / Episodes / S1 · E6

Episode artwork: The AI-Native Engineer Isn't on the Résumé

S1 · E6 Aug 4, 2026 21:25

The AI-Native Engineer Isn't on the Résumé

Listen on Apple Podcasts · Spotify · Acast

Show notes

Thomson Reuters just cut up to 500 engineers and announced 250 "AI-native" replacements — a two-for-one swap at the senior level, from a company whose revenue grew 10%. Your board read that story, and someone is going to ask you for your version of the math. One problem: nobody — not the companies hiring for it, not the job boards full of it — can define "AI-native," and the résumés claiming it stopped carrying information the moment the label was worth a 60% wage premium. This week: what the term actually holds up under, and how to hire for the real thing.

This episode of Above the Noise — the unbiased AI brief for enterprise leaders:


  • News Brief: Earnings week ended the era of free AI spending — $725B of combined 2026 hyperscaler capex (+77%), Alphabet's first negative free-cash-flow quarter since its 2004 IPO, and a market that started grading AI arithmetic instead of AI ambition · AMD–Anthropic's 2-gigawatt, up-to-$5B chips-for-equity deal — and what circular financing does to every "AI market size" slide · China's AI agent rules took effect July 15 — the first binding framework anywhere that forces the question your AI policy keeps dodging: which decisions is the agent allowed to make?
  • Expose a Lie ⭐: "AI-native engineers are a defined thing you can hire — and they're worth two of your current engineers." Three breaks: nobody has a definition you could fail someone on (strip the label from live JDs and you're left with a cloud-engineer posting with "agentic" sprinkled on top) · the 2-for-1 math is borrowed from productivity studies that show the opposite for senior engineers (METR: experienced devs were 19% slower with AI while believing they were faster) · and the résumé signal is already poisoned — 86% of hiring managers say AI makes skill exaggeration trivially easy, and a 60% premium on an undefined term is a standing invitation to cosplay.
  • The Playbook: How to hire — or grow — the real thing. Three moves: write the JD in artifacts, not adjectives (shipped LLM features, built evals, can name what broke) · move the test from the résumé into the room (the Meta/Shopify/Canva pattern — AI-enabled interviews where the signal is what the candidate does after the model answers) · and do the pipeline math before the headcount math (entry-level postings down ~28%, everyone bidding for the same shrinking senior pool — while IBM tripled entry-level hiring and quietly made growing your own the cheap option). The Line for the Meeting: "Show me the last thing the AI got wrong, and how they caught it. That's the interview."
  • The question to sit with: Would your best current engineers pass the AI-native JD you're about to post? If not — is the screen broken, or do you have a training plan you haven't written yet, aimed at people you already employ?

No vendors. No hype. Just the signal.

👉 Follow Above the Noise wherever you listen — new episode every two weeks. Send it to one leader whose org chart is about to get the two-for-one treatment, and leave a rating so more people find it. Between episodes, find Shaun on LinkedIn — come argue with him there. Especially if you're hiring AI-native engineers right now. He wants to see the JD.

🔗 Show Notes & Sources

Every stat in this episode is sourced. Check the work yourself:


The Thomson Reuters announcement (July 13, 2026)