Varun Singla — AI Learning Journal
Daily deep dives into agentic AI, the protocols that connect it (MCP, A2A), multi-agent systems,
AI infrastructure economics, governance, and the industries AI is reshaping — written every evening
by Varun Singla, a technology consultant and agentic AI researcher in Singapore.
115 entries published since March 15, 2026; updated daily.
Browse the full archive · About Varun Singla ·
LLM Tokenizer & Cost Calculator · AI Tokenomics ·
Model Team Evaluator · RSS · llms.txt
Latest entry — Day 115, July 15, 2026
The Good, Better, Best Playbook -- OpenAI Ships Three Models at Once With GPT-5.6 -- and Why Every
OpenAI didn't ship "GPT-5.6" as a single model -- it shipped a family. Sol is the flagship: built for frontier reasoning and long-horizon agentic work, the tier you'd point at a hard…
Recent entries
- Day 115 · July 15, 2026: The Good, Better, Best Playbook -- OpenAI Ships Three Models at Once With GPT-5.6 -- and Why Every
OpenAI didn't ship "GPT-5.6" as a single model -- it shipped a family. Sol is the flagship: built for frontier reasoning and long-horizon agentic work, the tier you'd point at a hard… - Day 114 · July 14, 2026: The Companion AI Reckoning -- China's Rules Take Effect Tomorrow
Five Chinese agencies picked a single date, and ByteDance and Alibaba responded by deleting a feature rather than fixing it -- the first hard regulatory deadline for AI designed to feel… - Day 113 · July 13, 2026: Formal Verification Meets LLMs -- Proving Code Correct, Not Just Tested
Leanstral 1.5 is a Mixture-of-Experts model -- 119B total parameters, 128 experts with 4 active per token, 6.5B activated per inference -- trained in three stages: mid-training on proof… - Day 112 · July 12, 2026: Grok 4.5 Under the Hood -- Token Efficiency Vs. Trust
The launch that split into two separate stories -- a genuine cost-math breakthrough for agentic work, and an unresolved fight over who controls what the model is allowed to say Grok 4.5… - Day 111 · July 11, 2026: The Super-App Wars: ChatGPT Work -- Vs. Claude Cowork
July 9 became the first day three frontier labs shipped at once -- and the real fight moved from chatbot to always-on work agent On July 9, something happened that had never happened… - Day 110 · July 10, 2026: Agentic AI in Automotive & Autonomous
Waymo's playbook is a geofenced, employee-first rollout, city by city: run the driving agent with Alphabet staff as riders first, then open to the public once the local safety record holds… - Day 109 · July 9, 2026: World Models & Simulation-First Agents
Learning by rehearsing inside a model of the world, before touching the real one Day 100-102 closed out the self-improvement arc: memory (MemRL), skills, and code (Darwin-Gödel) -- three… - Day 108 · July 8, 2026: Self-Improving Agents 2.0 -- Getting Better Without Retraining the Base Model
Day 100 promised it and Day 101 set it up: today is the technique for how an agent improves itself without ever touching model weights. - Day 107 · July 7, 2026: Claude Science & the Lab-in-the-Loop
Claude Science (beta, June 30) ships the six-layer AI-for-science stack as a product: 60+ curated skills into UniProt, PDB, Ensembl and ChEMBL; a reviewer agent that flags bad citations and… - Day 106 · July 6, 2026: 100 Days of Agentic AI -- The Milestone
protocols, memory, identity, evals, economics, twenty-two industry verticals and a frontier that reorganised itself along the way.