Know what's true and do what's right
Hint: managing someone is actually hard, and most times not worth it. Most people don’t use AI agents daily because delegation (to another human) is a learned skill. Using AI agents is actually a lot like doing natural language programming: specifying requirements, verifying outputs, guiding the flow, giving context. Also, using an AI agent requires Read More The post Why AI agents aren t adopted…
I’ve said this before and I’ll say it again: Using LLMs-written content professionally will spoil your career, especially if you’re in early stages. Everyone worth their salt that I know is hugely aversive to reading slop and despite your smart attempts, they can sense signatures of AI writing (even if they can’t articulate what gave Read More The post Write the first damn draft yourself appeared…
What can 🇮🇳 do to get ahead Here’s some unsolicited advice to the government on what to do to help boost India’s sovereign model story. 1. Mandate domestic models for all non-critical functions Government needs to be a guaranteed buyer for homegrown models. There are tons of applications that don’t require SOTA performance. (To be Read More The post Unsolicited advice to the Indian govt on…
Math is not what it used to be anymore I feel AI is breaking our usual, shared understandings of what words mean. For the first time ever, we have a new kind of cognitive labor that isn’t humans, and that’s breaking norms for what we’ve forever implicitly assumed to be universally true. All our assumptions about how Read More The post AI is challenging what words mean appeared first on Inverted…
The answer is in economics, not talent. The biggest hurdle for an Indian company trying to develop frontier model is that right from the get go, it has to compete with Anthropic and OpenAI free plans + all Chinese open weight models. Chinese models mostly flourished due to the demand from their domestic market which Read More The post Why are Indian companies not developing frontier models?…
Investors underwrite scaling laws because the alternative (having an Einstein discover better algorithms) is high variance What would it take to make progress towards general intelligence, where general stands for any problem that might arise in our world? Since our world is big and open-ended, the quest for general intelligence becomes a quest for solving Read More The post Why frontier labs are…
Saying that LLMs are just next token predictors is underselling these beasts to a mind numbing degree.First, LLMs aren’t just predicting the next token. They plan ahead the loss function is the average of cross entropy across all future tokens in a context window and the attention has access to all previous tokens. So, at a Read More The post LLM are universal simulators appeared first on Inverted…
Wigner wondered about the unreasonable effectiveness of mathematics in describing our universe. That is, why are our physical laws as described by precise mathematics produce predictions that turn out to be true? Let’s look at it from three perspectives. #1 We grasp the graspable bits By definition, we model aspects of reality that we are capable of Read More The post Why math works so well in…
Daniel Dennett, who recently passed away, is an inspiration for people like me who love grand ideas that connect many different fields at a deep level (and be rigorous about that; well as rigorous as grand theories can be). His thinking is difficult to categorize: part philosopher, part scientist, he had many good ideas on topics ranging from evolution Read More The post How mind bootstraps itself…
I just finished this tremendously readable and provocative book: Escape from Shadow Physics. The book is about foundations of quantum mechanics. To be precise, it’s about why we should not give up on trying to find a deeper theory of quantum mechanics. The usual story in QM is that the theory is complete and the probabilistic Read More The post Science often proceeds one grave at a time appeared…