> This post is 100% human-written. Claude was used for feedback and to assist with the linker symbol diagram. Cursor was used for feedback and to ensure examples were compilable.
Love this, I hope every blog have the same disclaimer about how AI is used.
It will pay for itself if you spend >300$ per month. I personally wish Waymo have a 399$ per month subscription that give 2 free ride per day so I don't need to own a car just for work.
An outage could cost Amazon ~millions to tens of millions. Most of the time, we want the junior to learn from the outage and fix the process. With AI agent, we can only update the agent.md and hope it will never happen again.
It interesting to see that the eval set becoming more and more expensive. Previously we just need to evaluate one test set, right now we need to create a lot of diffs and run a lot of tests.
I think the good thing about it is that if you are given good specification, you are likely to get good result. Writing a C compiler is not something new, but it will be great for all the porting projects.
I miss entering flow state when coding. When vibe coding, you are in constant interruption and only think very shallow. I never see anyone enter flow state when vibe coding.
I think It's not because AI working on "misaligned" goals. The user never specify the goal clearly enough for AI system to work.
However, I think producing detailed enough specification requires same or even larger amount of work than writing code. We write rough specification and clarify these during the process of coding. I think there are minimal effort required to produce these specification, AI will not help you speed up these effort.
According to https://cdn.realfood.gov/Daily%20Serving%20Sizes.pdf, their recommendations do not meet their calories goal. Eg, for 2000 calories, you can eat 4 egg, 3 cup of milk, 4 slice of bread, 2 apple and 3 tbsp of oil per day.
It seems O3 following trend of Chess engine that you can cut your search depth depends on state.
It's good for games with clear signal of success (Win/Lose for Chess, tests for programming). One of the blocker for AGI is we don't have clear evaluation for most of our tasks and we cannot verify them fast enough.
Love this, I hope every blog have the same disclaimer about how AI is used.