White Papers

Bitcoin: A Peer-to-Peer Electronic Cash System
Improving Language Understanding by Generative Pre-Training, the OpenAI GPT 1 paper. See also the blog post.
Language Models Are Unsupervised Multitask Learners, the OpenAI GPT 2 paper. Alternative link: OpenAI CDN
Language Models are Few-Shot Learners, the OpenAI GPT 3 paper.
Training language models to follow instructions with human feedback, the OpenAI InstructGPT paper. InstructGPT is, roughly speaking, ChatGPT. See ChatGPT announcement for details on how InstructGPT led to ChatGPT.
The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity, Apple's [in?]famous paper. See its announcement.

A newsletter for programmers

Yo! This is Taro. I've been doing JavaScript for years and TypeScript for years. I have experience with many programming languages, libraries, frameworks; both backend and frontend, and in a few company roles/positions.

I learned a few things over the years. Some took more effort than I wish they had. My goal with this blog and newsletter is to help frontend and backend developers by sharing what I learned in a friendlier, more accessible and thorough manner.

I write about cool and new JavaScript, TypeScript and CSS features, architecture, the human side of working in IT, my experience and software-related things I enjoy in general.

Subscribe to my newsletter to receive notifications when I publish new articles, as well as some newsletter-exclusive content.

No spam. Unsubscribe at any time. I'll never share your details with anyone. 1 email a week at most.

Success!
You have subscribed to Taro's newsletter Shoot!
The server blew up. I'll go get my fire extinguisher — please check back in 5.

Email Name