Stymied datacentre projects threaten global AI revolution
THE SO WHAT
Cancelled and challenged hyperscale projects—from Prince William Digital Gateway onward—show that power, water, and local politics are now the gating factors for AI growth, not GPUs. When you plan AI capacity, model permitting timelines and grid constraints with the same rigor you apply to model roadmaps.
READ THE SOURCE
MORE FROM THE WIRE
.png)
Deep & Emerging TechBritish Space Startup Launches Longevity Lab Into Orbit
Space-based biolabs feeding AI models on protein behavior tie longevity, AI, and space into a single R&D stack. If you’re in drug discovery or tools, assume your competitors will be training on data you literally cannot generate on Earth.
Deep & Emerging Tech'No single organisation can tackle it alone': VodafoneThree says its new security process blocked two million SMS fraud messages
VodafoneThree and Barclays cutting SMS fraud by 25% via shared intelligence shows telco–bank coalitions can actually move the fraud baseline. If you’re a comms or financial provider still running solo defenses, you’re effectively underwriting the attackers’ R&D.
Deep & Emerging TechSamsung’s Q2 profit jumps 19-fold to 89 trillion won on AI memory demand
A 19x profit jump to roughly 89.4 trillion won on AI memory demand underlines where the current value pool sits—in the components feeding model training and inference. If you’re upstream or downstream of HBM and memory, assume pricing and allocation risk will shape your roadmap more than model benchmarks this cycle.
Deep & Emerging TechExtreme SK Hynix Stock Swings Add Wild Card to $28 Billion Deal
SK Hynix’s volatility around a $28B US listing is a reminder that AI-chip exposure now brings hedge-fund-level swings to what used to be boring memory. Operators relying on these suppliers should diversify and lock in terms where possible—equity turbulence can spill into supply and pricing.