Andrew Moore, Google Cloud's new head of AI, on mistakes companies make in adopting AI, working with US military, and China's importance to Google's AI plans
Why SAP Is Buying Qualtrics for $8 Billion Tweets: Catherine Breslin / @catherinebuk : “AI is about using math to make machines make really good decisions. At the moment it has nothing to do with simulating real human intelligence.” http://www.technologyreview.com/ ... Subrahmanyam Kvj / @sub8u : How not to go about AI... Good read."I've got this massive amount of data—surely there's some value I can get out of it"http://www.technologyreview.com/ ... http://twitter.com/...
Context & Ripple Effects
Andrew Moore's first major interview as Google Cloud's head of AI lands one month after reporting on his military ties, including leading a task force alongside Robert Work, who reportedly created Project Maven — the program whose employee backlash made Google's defense posture a live question. The interview is effectively his answer: he defends working with the US military while staking out an enterprise-adoption message that 'I've got massive data, surely there's value in it' is the wrong way in.
First-order effects
- Enterprises evaluating Google Cloud get a named AI leader articulating a discipline-first adoption doctrine — start with a decision math can improve, not a data dump — which becomes sales collateral against AWS and Microsoft in cloud AI pitches.
- Moore's public case for military work repositions Google Cloud for government customers after Project Maven, directly answering the internal-ethics pressure documented when Google's proliferating AI made its ethical ramifications more consequential and murky.
Second-order effects
- Rival clouds must now field comparable AI leadership voices and government-facing narratives, since Moore's dual pitch — enterprise rigor plus defense legitimacy — sets the terms of competition for both budgets.
- The 'data first' critique pressures vendors selling raw-data platforms to repackage offerings around specific decisions and outcomes, shifting procurement conversations toward measurable use cases.
Third-order effects
- If the pattern holds, successful AI adoption concentrates in enterprise installations that automate decisions rather than simulate human intelligence — consistent with the later argument that AI's first productive wave resembles early computing's enterprise-led, job-cutting phase.
- Defense and intelligence work becomes a structural pillar of cloud AI strategy, forcing every major provider to reconcile national-security revenue with workforce ethics — a tension Google was still navigating years later under Thomas Kurian's Gemini Enterprise push (Kurian Q&A).
The trend: Cloud providers are institutionalizing AI under named executives who must simultaneously win enterprise adopters, court government buyers, and manage an ethics backlash — with China's AI ambitions sharpening all three pressures.