63 long-form pieces on architecture, training, inference, agents and safety.
Replace the reward model with a function that simply checks the answer, and a frontier reasoning model falls out of pure reinforcement learning. The catch is what 'checkable' quietly assumes, and what the model learns to exploit.
DPO is the right default for almost every preference-tuning project in 2026, but the cases where PPO still wins are sharper and more common than the simplicity pitch admits.
Teams tune the embedding model and the vector database, then wonder why answers are still wrong. Both are the least important parts of the stack. Retrieval is a two-stage ranking problem, and information retrieval solved the shap…
Test-time compute makes models smarter but slower and pricier. Sleep-time compute moves the thinking to before the user asks, cutting the compute needed for the same accuracy by roughly 5x and the cost per query by 2.5x when cont…
Autoregressive decoding forces even the largest language models to emit one token at a time, bottlenecked by memory bandwidth rather than compute. Speculative decoding breaks this sequential constraint by letting a lightweight dr…
Most teams reach for fine-tuning when they should reach for retrieval. The cost is not just dollars - it is months of confused engineers trying to figure out why the model started lying again.
Language models generate text one token at a time by sampling from a probability distribution over their entire vocabulary. Constrained decoding intervenes at that sampling step, masking out every token that would violate a targe…
A 7B model that thinks for thirty seconds can outscore a 70B model that answers instantly. Test-time compute is the lever that makes this possible, and it is quietly rewriting the economics of model design.
DeepSeek shipped a 1.6T-parameter model in April 2026 with native Ascend inference support while OpenAI committed $500B to Stargate. Both can be right, but only one of them is a moat.
A frontier model can write a function. An agentic runtime can read a codebase, plan a migration, edit forty files, run the tests, fix what broke, and open a pull request. The difference is not intelligence; it is the system that …
Most teams still treat Anthropic as 'the Claude API.' That framing misses the platform that has grown around it: managed agents, a universal integration protocol adopted by every major AI vendor, a governance layer with 28 securi…
Anthropic stopped shipping a chatbot and started shipping an ecosystem. From Artifacts to Cowork to Design, the through-line is the same: move the model out of the chat box and into the work. For marketing teams, that change is n…