Stories by dbreunig

FLUX.1-Krea and the Rise of Opinionated Models

Does the Bitter Lesson Have Limits?

Moonshot Used RL for Qualitative Tasks to Write Better

Kimi 2 Became One of the Best Tool-Using Models

Rewriting Training Data Improved Kimi 2's Performance

Task Delegation by Experts Is the AI Metric That Matters

Ask HN: How are startups dealing with key talent poaching?

Techniques for avoiding common long context fails

The Ways Long Contexts Fail

An Agentic Case Study: The Messy Work of Building a Gemini Agent to Play Pokémon

The drawbridges come up: the dream of a interconnected context ecosystem is over

Let the Model Write the Prompt: Using DSPy to Decouple Your Task from the LLM

Comparing Claude System Prompts Reveal Anthropic's Priorities

Using 'Slop Forensics' to Determine Model Lineage

Chatbots Trigger Guardrails Differently Based on Your Age, Gender, and Fandom

Breaking Down Claude's 26k+ Token System Prompt

DuckDB is probably the most important geospatial software of the last decade

Explaining Reasoning Models: How they work, how they're built, & why they matter

AI-Powered Apps Are Changing the Dynamic Between Domain Experts and Developers

AI Chatbots Are Like Observational Comics

1
2
3
4
5
6
7
8
9
10