简体中文
EnglishWhat Kind of Writing and Topics Are on Jet New (jetnew.io)?
Updated
Jet New's site is a personal notes site rather than a tutorial library or news outlet. Its writing is organized into two sections, UPCOMING and ARCHIVE, and the author states plainly that "all writing are continuous works-in-progress." If you want polished, finished guides, this is not the place; if you want to see how one person thinks through AI research and building, the archive gives you a readable sample.
How the site is organized
The homepage carries a short description—"Notes on what I'm thinking about, learning, and working on"—and then splits content into two lists:
| Section | What appears there | Example |
|---|---|---|
| UPCOMING | Announced but not yet published pieces | "Paper, Wood, Leather - Why I'm building Atlas, an AI tool for thought" (dated Oct 2026) |
| ARCHIVE | Published notes, listed with dates | "100 Lessons from 1 Year of AI Research" (Dec 2021) |
The archive entries are dated, so you can read them as a rough timeline of the author's interests rather than as a curated course.
What topics the archive covers
The visible archive points to three overlapping areas:
- AI research experience — "100 Lessons from 1 Year of AI Research" (Jul 2021 is the reinforcement learning piece; the lessons piece is Dec 2021) reads as reflective, lesson-style writing drawn from doing research.
- Getting started with a technical subject — "Getting Started with Reinforcement Learning" (Jul 2021) is the closest thing to an entry-point piece, though it sits within a personal notes frame.
- Running technical workshops — "Why You Should Conduct a Technical Workshop" (Mar 2021) is advice-oriented but grounded in the author's own practice.
The site also identifies the author's current work: he says he is "building Atlas, a tool for understanding research," and the page lists Atlas (an AI research workspace), AgentScale AI, and AI software consultancy as related projects. That context explains why the notes lean toward research tooling and applied AI rather than general programming.
What to expect from the writing itself
Two traits shape how you should read it:
- Work-in-progress by design. The author explicitly frames every piece as continuously revised. Treat claims as thinking-in-public, not settled reference material.
- Notes, not tutorials or news. There are no step-by-step menus, no product reviews, and no dated news coverage. The value is in the observations and the framing of problems.
For example, if you are weighing whether to run an internal workshop on a technical topic, the workshop piece is the kind of thing to skim for the author's reasoning and experience—then adapt it to your own team rather than following a fixed procedure.
Who it suits
- Readers curious about Atlas and the thinking behind an AI tool for research will find the UPCOMING piece directly relevant once published.
- People starting in reinforcement learning can use the getting-started note as one practitioner's orientation, alongside more structured courses.
- Anyone expecting a maintained blog with regular cadence should note the sparse, irregular archive dates (2021 entries, then a 2026 upcoming piece).
If your goal is a complete curriculum or current AI news, look elsewhere; if your goal is a compact view of one builder's research notes, the archive is short enough to read in one sitting.