Now Playing
Huawei's Smuggled HBM, Microsoft's $170B Bet & Google's School AI Lock-In
Audio is available on Spreaker — see link below.
Smugglers are melting solder off Korean memory chips to feed Huawei's AI accelerators — and that's just the start. Today's briefing covers Microsoft's $170B capex commitment, China's full-lifecycle AI regulations, and Google forcing schools into an AI ultimatum.
Audio is available on Spreaker — see link below.
Smugglers are melting solder off Korean memory chips, repackaging them as low-end server components, and shipping them straight into Huawei's AI supply chain. That's not a theoretical vulnerability in US export controls.
The smuggling story connects directly to a broader pattern. China's reliance on these workarounds signals that domestic chip production still can't meet demand for frontier AI development.
Meanwhile, the infrastructure economics of AI just got a hard number attached to them. Microsoft issued a twenty-five billion dollar bond, its first since twenty-seventeen, to fund accelerating capital expenditure.
That infrastructure logic is also behind this week's Lumen-Microsoft announcement. The two companies deepened their partnership to connect distributed data centers, office environments, and cloud deployments for enterprise AI workloads.
A different kind of pressure is being applied in education. Google rolled out Gemini Guided Learning as the default in Google Classroom and replaced AI Overviews in Google Search for school accounts.
One more story worth tracking. China's Foreign Ministry rejected a report alleging that Beijing had orchestrated intelligence gathering on Nvidia CEO Jensen Huang through a Singapore data center executive.
The through-line across all of this is execution risk. China's smuggling networks are under pressure as enforcement tightens and competition among dealers increases, but domestic HBM supply isn't ready to fill the gap.
Chapter summary auto-generated from the verified script. Listen to the full episode for the complete content.