Trump administration reaches a trade deal to lower Taiwan’s tariff barriers
https://apnews.com/article/trump-taiwan-china-trade-deal-2b1743397ba33010463d41132b75ce53
This so-called “trade deal” is less about economic pragmatism and more about Washington doubling down on geopolitical brinkmanship with China. Lowering Taiwan’s tariff barriers is framed as a win-win for free trade, but the real currency here is leverage—fuelling Taiwan’s tech sector to further antagonize Beijing. Investors should be wary: this deal risks escalating supply chain decoupling, increasing uncertainty for semiconductor supply and global markets far beyond Taiwan’s shores.
OpenAI sidesteps Nvidia with unusually fast coding model on plate-sized chips
https://arstechnica.com/ai/2026/02/openai-sidesteps-nvidia-with-unusually-fast-coding-model-on-plate-sized-chips/
OpenAI’s pivot away from Nvidia dependency is hailed as innovation, but it exposes cracks in the current AI hardware monopoly. Plate-sized chips may boost speed, but such radical hardware shifts risk fragmenting AI development standards and creating new bottlenecks. The race to replace Nvidia is less a technological leap and more a frantic scramble to diversify chip supply amid Washington’s semiconductor restrictions against China, inadvertently fueling hardware arms races that will destabilize the market.
Indian conglomerate Adani Group announces plans to invest $100B to build renewable energy-powered AI-ready data centers across India by 2035
https://www.reuters.com/business/energy/indias-adani-invest-100-billion-ai-data-centres-by-2035-2026-02-17/
Adani’s massive $100 billion bet on AI-ready data centers is framed as sustainability meets tech ambition, but the real story is India’s desperate bid to carve out sovereignty in AI infrastructure before the US-China tech war locks out emerging players. Such gargantuan investments carry political and execution risks, especially under India’s volatile regulatory environment. This is not just about green AI; it’s a strategic hedge against being a peripheral supplier to Western and Chinese AI giants.
UK activist investor Palliser says Toto, Japan’s largest toilet maker, is an “undervalued and overlooked” AI play due to its “advanced ceramics” used in chips
https://www.ft.com/content/4252e45f-75fb-4dfc-aebe-72de48b7fb8e
It’s easy to scoff at a toilet maker being pitched as an AI semiconductor play, but Palliser’s call highlights a deeper undervalued node in supply chains: advanced ceramics critical to next-gen chips. Investors ignoring these “hidden tech” suppliers risk missing the looming squeeze from material shortages and geopolitical controls on rare materials. Toto’s pivot may seem quirky, but it signals the fracturing of traditional supply chains into obscure specialty niches that could dictate the semiconductor race.
Sources: SpaceX and xAI are competing in a $100M DoD contest to make voice-controlled, autonomous drone swarms; OpenAI is helping Applied Intuition’s submission
https://www.bloomberg.com/news/articles/2026-02-16/spacex-to-compete-in-pentagon-contest-for-autonomous-drone-tech
Behind the flashy headlines about autonomous drone swarms lies a disturbing militarization of AI led by private firms like SpaceX and OpenAI. This $100 million DoD contest is not just innovation—it’s a harbinger of AI-powered battlefield asymmetries that will reconfigure global security paradigms. The private sector’s deepening role in defense tech blurs lines between commercial AI ethics and lethal autonomy, raising urgent questions that mainstream coverage is too soft to address.
India eyes $200B in data center investments as it ramps up its AI hub ambitions
https://apnews.com/article/india-modi-vaishnaw-ai-investments-4e170da0a3b883a9659569dd538e9019
India’s $200 billion data center push is portrayed as a bold step to become a global AI hub, but the rush exposes a broader structural vulnerability: overreliance on massive infrastructure building without guaranteed access to cutting-edge semiconductor technology. India’s ambitions will bump against entrenched US-China export controls and supply chain bottlenecks. Without parallel moves in chip manufacturing and talent retention, these investments risk creating stranded assets rather than genuine tech sovereignty.
Sources: Hacker News, Techmeme, AP News, Ars Technica | Compiled 2026-02-17