If you are good at code review, you will be good at using AI agents
Read it on seangoedecke.com Using AI agents correctly is a process of reviewing code. If you’re good at reviewing code, you’ll be good at using tools like...
The whole point of OpenAI's Responses API is to help them hide reasoning traces
Read it on seangoedecke.com About six months ago, OpenAI released their Responses API, which replaced their previous /chat/completions API for inference. The...
'Make invalid states representable' considered harmful
Read it on seangoedecke.com One of the most controversial things I believe about good software design is that your code should be more flexible than your...
An unofficial FAQ for Stripe's new "Tempo" blockchain
Read it on seangoedecke.com Stripe just announced Tempo, a “L1 blockchain” for “stablecoin payments”. What does any of this mean? In 2021, I was interested...
Seeing like a software company
Read it on seangoedecke.com The big idea of James C. Scott’s Seeing Like A State can be expressed in three points: Modern organizations exert control by...
Do the simplest thing that could possibly work
Read it on seangoedecke.com When designing software systems, do the simplest thing that could possibly work. It’s surprising how far you can take this piece...
Finding the low-hanging fruit in tech companies
Read it on seangoedecke.com Suppose your job is to pick fruit in a giant orchard. The orchard covers several hills and valleys, and is big enough that you’d...
Everything I know about good API design
Read it on seangoedecke.com Most of what modern software engineers do1 involves APIs: public interfaces for communicating with a program, like this one from...
Don't feed me AI slop
Read it on seangoedecke.com In the early days of any new technology, the relevant social norms are still being workshopped. For mobile phones, that meant...
The famous "bottomless pit" AI greentext is fake
Read it on seangoedecke.com Many people believe1 this is the best piece of art or humour that AI has ever produced: This was generated three years ago by...
Is chain-of-thought AI reasoning a mirage?
Read it on seangoedecke.com Reading research papers and articles about chain-of-thought reasoning1 makes me frustrated. There are many interesting questions...
What's the strongest model you can train on a laptop in five minutes?
Read it on seangoedecke.com What’s the strongest model I can train on my MacBook Pro in five minutes? I’ll give the answer upfront: the best 5-minute model I...
OpenAI's new open-source model is basically Phi-5
Read it on seangoedecke.com OpenAI just released its first ever open-source1 large language models, called gpt-oss-120b and gpt-oss-20b. You can talk to them...
AI interpretability has the same problems as philosophy of mind
Read it on seangoedecke.com If you want to know how a particular AI language model works, the current state-of-the-art approach is to use a “probe”. Pick one...
Why I don't allow AI-generated content on my blog
Read it on seangoedecke.com I will never include any AI-generated content on this website. In general I’m a big AI believer, and I use large language models...
Trying to impress engineers, trying to impress managers
Read it on seangoedecke.com In the first few years of my career, I knew next to nothing about how to do good work in a tech company. The senior and staff...
The refusal problem in AI language models
Read it on seangoedecke.com Everyone’s interested in what language models can do. But what about what they will do? Try giving ChatGPT 4o a menial task that...
Can small AI models think as well as large ones?
Read it on seangoedecke.com An AI trend that’s emerged in the last few months1 is the idea of a “cognitive core”. Instead of trying to build the largest,...
What kind of work I want (in 2025)
Read it on seangoedecke.com In 2021, I wrote this post describing the kind of work I wanted then. Almost five years later, it’s time for an update. What kind...
Do not yell at the language model
Read it on seangoedecke.com A few days ago, the news went viral that Replit’s AI coding agent went rogue and deleted a production database. A lot has been...
