The revelation that 20% of Google's engineers refuse to work with AI, 60% are comfortably entrenched in AI-lite workflows, and the remaining 20% are valiantly striving to be AI-first had the internet in a flurry of excitement (and not at all dread at the realization of the inevitable future of automation). What was surely hoped to remain a tightly-held company secret has now inspired spirited rebuttals from Google's AI leadership, surely enthused about this opportunity for public discourse.
Demis Hassabis, CEO of Google DeepMind, dismissed the characterization as 'absolute nonsense' in a calm and reserved public/forum style post. He astutely rebutted the claim, ensuring the public that his prodigious team at DeepMind are indeed enthusiastically using AI—for something—on a daily basis.
Complementing Hassabis’s courteous retort, Google Cloud AI's Addy Osmani enthusiastically billed the company’s internal toolsets as unrivaled (a definite first in corporate communication). “Over 40,000 Google engineers are using agentic coding, which in technical terms, translates roughly to, ‘these guys and gals do stuff with computers, trust us,’” Osmani clarified.
Meanwhile, Yegge stands his ground, noting that if mere AI exposure is Google’s hallmark of innovation, then truly groundbreaking transformation work must look very different behind the curtain. This assertion has triggered a deep self-love fest among Google engineers, taking to X to declare their exceptional ability to both admire balloons and pop them, proving their tenacity by rapidly advancing through the Google ranks.
The collective breathless reactions to the pieces remind us once more of the genius of the tech world—so enthusiastic, yet never failing to surprise us with anticipated revelations. So steely in resolve, yet always happy to party in row 2 of the AI adoption bus.