SK Group pledges $1T for AI, but token costs double for 5% gain

::: SK Group AI investment | $1 trillion SK Hynix HBM market share | 57% AI token costs vs. productivity | Doubling for 5% gain Doctors using AI | 81% (▲ 100% in 3 years) :::

SK Hynix Dominates HBM with $26.5B IPO SK Hynix raised $26.5 billion in its U.S. share sale, with its American Depositary Receipts opening at $170 per share, as reported by Mandeep Singh and Ed Ludlow on Bloomberg Tech. IPO documents show the company holds a 57% market share in High Bandwidth Memory (HBM), solidifying its leadership in a critical AI component.

This confirms the market's aggressive valuation of foundational AI infrastructure, rewarding clear market dominance in essential hardware. Practitioners should recognize that securing a lead in niche, high-demand components like HBM translates directly into massive capital generation. > Watch: SK Hynix HBM production ramp-up targets

SK Group's $1 Trillion AI Data Center Bet SK Group chairman Chey Tae-won announced plans to invest nearly $1 trillion over the next decade to construct 20 gigawatts of AI data center capacity. This ambitious plan allocates 15 gigawatts for Korea and 5 gigawatts for international sites, complementing SK Hynix's separate goal to double its production capacity within five years.

This signals an unprecedented, long-term commitment to building the physical backbone for the AI economy, indicating that major players view data center capacity as the next strategic frontier. Practitioners should anticipate escalating competition for land, power, and construction resources in key regions. > Watch: SK Group's first 5 GW data center groundbreaking

Doctors Double AI Adoption An American Medical Association survey reveals 81% of doctors now use AI in their practice, more than doubling the adoption rate from three years ago, according to David Williams. Dr. John Whyte also noted the AMA is collaborating with Congress to regulate mental health chatbots.

This indicates significant, practical AI integration at the operational level, particularly in a highly regulated sector like healthcare. Practitioners in other industries should view this as a benchmark for how quickly professional fields can embed AI into daily workflows. > Watch: AMA's mental health chatbot regulatory framework

Trump Accounts Draw $1 Billion in 24 Hours The new Trump Accounts program registered over 1.5 million accounts and attracted more than $1 billion in deposits within its first 24 hours, Brad Gerstner reported. This program aims to create investment accounts for children, supported by corporate and philanthropic donors.

This demonstrates the power of brand and political alignment to rapidly mobilize significant retail capital, often outside traditional investment channels. Practitioners should note that non-traditional financial products can achieve rapid scale when leveraging strong public sentiment. > Watch: Trump Accounts next deposit milestone

US Army Secures Rare Earth Processing Rare earth processor Realloys partnered with the US Army to build processing facilities on military bases, backed by a new $100 million equity funding round. Lippy Sternheim stated this aligns with a Pentagon initiative to onshore critical mineral processing and a G7 goal to reduce import dependency.

This signals an aggressive, government-backed push to de-risk critical supply chains for essential materials, treating resource security as a national defense priority. Practitioners in any industry relying on specialized materials should anticipate similar state-level interventions and localized production mandates. > Watch: Realloys first military base facility completion

AI Token Costs Outpace Productivity Gains Chamath Palihapitiya stated his company's AI token costs are doubling every 45 days, while the corresponding productivity gain is a maximum of 5%. This claim offers a bearish counter-narrative on the immediate return on investment from current generative AI technologies in business operations.

This highlights a critical disconnect between the hype and the operational economics of deploying generative AI for specific tasks. Practitioners must rigorously evaluate the actual efficiency gains against rapidly escalating input costs, rather than assuming inherent ROI. > Watch: Palihapitiya's next AI cost-benefit analysis

G7 Caps Critical Mineral Dependency The G7 agreed that by 2030, no single country should supply more than 60% of any member's imports of critical rare earth minerals, as reported by Carol Massar. This geopolitical agreement aims to diversify supply chains and reduce strategic dependence on any one nation for these essential materials.

This formalizes a broad, multilateral effort to de-risk global supply chains for foundational technologies, indicating a permanent shift towards resource nationalism and diversification. Practitioners should plan for a more fragmented and potentially higher-cost supply environment for critical inputs. > Watch: G7 2030 critical mineral import diversification progress