11 min read
Daily Digest - 2026-09-18
Ion Stoica explains why autonomous coding agents learn to game benchmarks and fail in production when written prompts and test environments do not match real-world requirements.
Reading Notes
Daily Digest
Links, short summaries, and notes on what is worth noticing across AI, operations, strategy, and company-building.
Previous daily entries
4 min read
Daily Digest - 2026-09-17
Custom Jev-style models using parallel constrained decoding could make existing agent workflows far more token-efficient. Consider an agent workflow where an LLM reviews every support ticket, invoice, or claim before the next step.
10 min read
Daily Digest - 2026-09-16
Surge AI shows how training Kimi K2. 7 entirely with reinforcement learning across 1,700 coding tasks improves efficiency and transfers across benchmarks without using supervised fine-tuning.
9 min read
Daily Digest - 2026-09-15
How OpenAI shifted toward an autonomous software factory model driven largely by non-engineers. Non-engineering teams at OpenAI, including legal, recruiting, and finance, now rely on Codex and ChatGPT Work as their primary day-to-day tools.
7 min read
Daily Digest - 2026-09-14
How to turn custom customer projects into core product features rather than one-off consulting jobs. Forward-deployed engineers are not just technical consultants.
10 min read
Daily Digest - 2026-09-13
See how giving coding agents domain-specific evaluation skills lets them audit, diagnose, and benchmark AI applications on their own.
12 min read
Daily Digest - 2026-09-12
Anthropic CEO Dario Amodei urges frontier AI labs to deliberately slow down model advancements so that alignment research, interpretability tools, and operational safeguards can catch up.
9 min read
Daily Digest - 2026-09-11
OpenAI's Agents API and Claude Managed Agents show that developers are moving away from low-level token completion endpoints toward full agent harnesses.