Everything I write about artificial intelligence lands here. Less of the demo-day theater, more of what sits underneath it: the compute, the real economics, and the distance between what gets announced and what actually ships.
Google shipped Gemini 2.0, AMD shipped the MI300X with 180GB HBM3, Apple shipped the M4 Neural Engine. Three bets on where the AI stack…
Google just released Gemini 2.0, and the question nobody is asking is the right one. The model is 20% better at reasoning benchmarks than…
McDonald's runs one of the most used consumer AI platforms on the planet from its Chicago HQ, and the city should claim it as…
Chicago's AI ecosystem: world-class research from UChicago, TTIC, Argonne, and Fermilab; enterprise-focused applied AI in healthtech and logistics; talent costs well below coastal hubs.…
The comparison between Apple's iOS 18 with Apple Intelligence and Qualcomm's Snapdragon 8 Elite highlights a key architectural difference. Apple's system relies on cloud…
The Snapdragon 8 Gen 4 is not a mobile chip refresh. Qualcomm brings custom Oryon CPU architecture to phones for the first time, a…
Days apart, Google ordered 500MW of small modular reactors from Kairos and Amazon backed 5GW of X-energy SMRs. Big Tech has decided nuclear is…
Apple's Vision Pro ships at $3,499 and amazes without converting the mainstream, then Meta shows Orion, true AR glasses that split XR's future in…
Meta's Llama 3.2 is a multimodal model that runs locally on mobile devices or in your own data center with no API dependency. The…
Microsoft signed a 20-year deal to restart Three Mile Island's Unit 1 reactor and power its AI data centers. The clearest sign yet that…
The EU blocked Apple's on-device Siri AI from launching in Europe. Apple's architecture was specifically designed to keep data local. The EU blocked it…
MIT's breakthrough in AI reasoning capabilities, Intel's 18A process node for AI workloads, and generative AI transforming drug discovery, three developments that signal a…
Two weeks after Apple Intelligence and six days after the first Copilot+ laptops, Qualcomm's on-device AI bet spans phone and PC at once. The…
OpenAI makes GPT-4 Turbo with Vision generally available while AMD launches the MI300X AI accelerator, expanding options for developers building multimodal AI applications and…
GPT-4 Turbo ships measurably better reasoning. AMD MI300X brings 5.3 TB/s of memory bandwidth. Intel Xeon 6 argues the CPU still matters. Together they…
One maker's tech digest, weekly. Chips, AI, gaming, EVs — no spam.