Jesus Rodriguez

DeepMind's Genie represents a breakthrough in actionable world models, moving beyond passive video generation to create interactive, controllable AI environments where systems can understand agency and causality.

“Author of "The Sequence Knowledge #817: DeepMind Genie and Interactive World Models" in TheSequence”

The U.S.-China competition for AI chips dominates geopolitical strategy, with NVIDIA's dominance, Huawei's rise, and Taiwan's critical role in chip manufacturing creating a semiconductor cold war that extends beyond mainstream media coverage.

“Author of "The Sequence Opinion #714: The AI Chip Cold War: NVIDIA, Intel, Huawei and" in Substack”

Alibaba released Qwen-Max, the first trillion-parameter MoE model positioned as a flagship production offering with OpenAI-compatible API, enabling developers to deploy frontier-class models with minimal integration friction.

“Author of "The Sequence Radar #715: Qwen-Max: The Trillion-Parameter MoE You Can Actually Ship" in TheSequence”

AI inference cloud startups like Together AI and Fireworks AI are rapidly proliferating to meet demand for cost-effective model deployment, but the market faces intense competition from hyperscale cloud providers and GPU vendors, likely leading to commoditization where only top performers survive.

“Author of "The Sequence Opinion #710: The Inference Cloud Wars: Speed, Scale, and the Road to Commoditization" in TheSequence”

Hermes 4 successfully combines long-form reasoning with crisp assistant behavior through explicit thinking baked into training, while providing transparent reproducible methodology via DataForge and Atropos.

“Author of "The Sequence AI of the Week #713: Inside the Amazing Hermes 4, an Open Reasoning Model" in TheSequence”

“Author of "The Sequence #706: Tiny, Long, and Quantized: A Deep Dive into Gemma 3 270M" in The Sequence”