How AI companies spent billions rediscovering something TCS / Infosys / Accenture / Thought works  figured out decades back — and why the pattern was always going to repeat. #ForwardDeployedEnginner ashkrit.blogspot.com/2026/05…

From git clone to llm clone llm clone nextjs.org ashkrit.blogspot.com/2026/05…

Building on Rented Ground On April 22, 2026, Anthropic changed a checkbox on a pricing page. No announcement. No email. No deprecation notice. Just a quiet edit — and overnight, Claude Code disappeared from the $20/month Pro plan. ashkrit.blogspot.com/2026/05…

Stop Reaching out to Agents Industry keeps pushing toward the flashiest pattern on the menu, and teams keep mistaking complexity for capability. The truth is boringly simple: you should have to be forced up the ladder, not invited. #AgenticPatterns ashkrit.blogspot.com/2026/04…

Jeff Bezos had a simple heuristic for team size: if two pizzas can't feed the team, the team is too big. It was never really about pizza. ashkrit.blogspot.com/2026/04… #AgenticAgile

Great Claude Code Leak and Under the hood - Claude code | Codex | Gemini Every system prompt is a Natural Language Program,  list of instructions — a code of how an AI agent becomes reliable. ashkrit.blogspot.com/2026/04… #ClaudeCode #PromptDesign

AI Engineering Terms You Will Memorize and then forget LLM era has a reliable product cycle: someone coins a term, someone more famous endorses it, the internet credits the endorser, and LinkedIn does the rest. ashkrit.blogspot.com/2026/04…

Similarity is not Relevance LLMs are similarity machines. Every layer — retrieval, generation, alignment — optimizes for proximity. But similarity is geometric. Relevance is teleological. One asks: how close? The other asks: does this help? ashkrit.blogspot.com/2026/03…

Revisiting Classic paper - tail at scale ashkrit.blogspot.com/2026/03…

Why the AI your company just mandated was designed for a boardroom slide, not your workflow — and how the whole machine keeps spinning. ashkrit.blogspot.com/2026/03…

New Blog Post on - Say Goodbye to Boilerplate: Annotating Your Way to Powerful API Clients ashkrit.blogspot.com/2024/03…

Blog post on Vector database ashkrit.blogspot.com/2024/02…

Blog post of Data structure that supports access pattern of Map Sorted by X attributes. ashkrit.blogspot.com/2023/08…

Amazing! A 17-year-old teenager created this while striving to understand how applications are executed: cpu.land The book explains: * How CPUs execute your app's instructions * How the OS uses interrupts to switch CPUs between tasks * The way virtual memory allocates physical memory among processes * The processes of app compilation and linkage

7) "Layered Money" by @timevalueofbtc Bhatia offers an innovative perspective on #Bitcoin 's potential to become the first layer in the future of finance. It's a thought-provoking exploration of Bitcoin's role in the financial ecosystem.

Biases that distort our Decision Making: 1. Confirmation Bias - We interpret new information as confirmation of our existing beliefs. 2. Availability Bias - We tend to rely on information that comes to our mind easily/the quickest. 3. Action Bias - We favor action over inaction. That's why we sell or buy prematurely. 4. Overconfidence - We overestimate our own knowledge and ability. Often because we know too little to know better. (Less knowledge => more confidence) 5. Survivorship Bias - This is a sample bias that occurs when we assess only successful outcomes and disregard failures. What Biases do you struggle the most with?

It took me several hours to compile everything. I hope you'll like the PDF as it's worth more than thousands of $ in courses. Grab everything from Michael Mauboussin for free here: eepurl.com/h9kw29

10 highly underrated investing books (that will 10x your investing skills):

The CEO of OpenAI who helped create chatGPT: Sam Altman. He has a net worth of $250,000,000 from his businesses. I've spent 100s hours researching his blog and interviews. Here're 10 powerful lessons on how to be a great entrepreneur: