seangoedecke.com RSS feed

The whole point of OpenAI's Responses API is to help them hide reasoning traces

\'Make invalid states representable\' considered harmful

An unofficial FAQ for Stripe's new "Tempo" blockchain

Seeing like a software company

Do the simplest thing that could possibly work

Finding the low-hanging fruit

Everything I know about good API design

Don't feed me AI slop

The famous "bottomless pit" AI greentext is fake

Is chain-of-thought AI reasoning a mirage?

What's the strongest AI model you can train on a laptop in five minutes?

OpenAI's new open-source model is basically Phi-5

AI interpretability has the same problems as philosophy of mind

Why I don't allow AI-generated content on my blog

Trying to impress people you don't respect

The refusal problem in large language models

Can small AI models think as well as large ones?

What kind of work I want (in 2025)

Do not yell at the language model

The case for regulating AI companions

How to actually use Model Context Protocol

Vibe coding and the silent AI war inside tech companies

Practical notes on getting LLMs to generate new ideas

The three great virtues of an AI-assisted programmer

Building an RTS where you issue orders to AI agents

METR's AI productivity study is really good

Why we should anthropomorphize LLMs

Mecha-Hitler, Grok, and why it's so hard to give LLMs the right personality

The more senior engineers get, the more results matter

How I use LLMs to learn new subjects

Building tiny AI tools for developer productivity

How to become passionate about delivering shareholder value

What happens when engineers work more than one job

Continuous AI in software engineering