Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Astra for Law(openai.com ↗)
    361comments
  2. Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint(prismml.com ↗)
    75comments
  3. Goose: 1.16x faster than C++ and 1.12x than safe Rust, while memory safe(github.com/aardappel ↗)
    24comments
  4. Bend – A language that blocks AI mistakes via proof, on CPU and GPU(bend-lang.com ↗)
    153comments
  5. Hister: A private search engine for the pages you visit and the files you keep(github.com/asciimoo ↗)
    137comments
  6. Wax motor(wikipedia.org ↗)
    50comments
  7. Fujitsu launches made-in-Japan next-generation CPU FUJITSU-MONAKA(global.fujitsu ↗)
    198comments
  8. Alibaba releases Qwen 3.8 Omni Flash(qwen.ai ↗)
    5comments
  9. Telstra outage: The night a network decided the year was 2006(netnod.se ↗)
    2comments
  10. Flet 1.0 – Build cross-platform apps in Python(flet.dev ↗)
    34comments
  11. Diplodocus, Long Thought Exclusively American, Turns Up in Spain(sci.news ↗)
    20comments
  12. More than 100k people in Japan are now aged 100 or older(bbc.com ↗)
    122comments
  13. I Put Nam A2-Lite Inside an iRig HD X(playtaurus.com ↗)
    1comments
  14. CrowdSec Source Code Leak(crowdsec.net ↗)
    40comments
  15. Infinite-Parameter LLMs: Generating and Adapting Weights from Live Data(arxiv.org ↗)
    34comments
  16. How Uber Protects Against Retry Storms(uber.com ↗)
    23comments
  17. Why I didn’t sign the Fields medallists’ letter(gowers.wordpress.com ↗)
    313comments
  18. The most important product decision is what you don't build(liamnugent.me ↗)
    19comments
  19. Rate limits on GitLab.com are changing(about.gitlab.com ↗)
    106comments
  20. How do we prevent mathemathics from devolving into the Medieval Era of secrecy?(mathoverflow.net ↗)
    56comments
  21. CCC invites all model citizens to 40C3(ccc.de ↗)
    180comments
  22. TSMC revealing details about next gen A14 node(mapyourshow.com ↗)
    36comments
  23. The American Religion of Self-Storage Facilities(newyorker.com ↗)
    343comments
  24. Landing the Space Shuttle – A Flying Machine and the Thrill of a Lifetime(eaa.org ↗)
    5comments
  25. Zettascale (YC S24) Is Hiring ASIC/FPGA Engineers to Build Chips for ASI(zscc.ai ↗)
    discuss
  26. Show HN: Snapdrop: Instantly share files between devices. No setup, no signup(snapdrop.me ↗)
    19comments
  27. Running Ubuntu on the Lenovo IdeaPad Duet(vhaudiquet.fr ↗)
    25comments
  28. Computer Reset, Dallas(dfarq.homeip.net ↗)
    2comments
  29. Launch HN: Skillsync (YC W26) – AI chat sessions made portable across agents
    50comments
  30. Show HN: Share your AI Setup, Learn from others(mysetup.ai ↗)
    103comments

Show HN: Direction – a 4-week course for people afraid of shipping AI slop

3 pointsby 5mo agogivedirection.com
0 comments
https://www.givedirection.com/

Every week on HN there's a valid thread about vibe coding producing unmaintainable garbage, or someone who built something with Claude over a weekend and can't explain how it works or fix it when it breaks. The gap I keep seeing is that people can't define their problem, don't scope their work, and rarely have exit conditions to test against such that they avoid these. This set of steps is arguably the most important thing you can do if you want something (software or otherwise) that lasts.

Direction runs monthly cohort-based courses where small groups go from idea to shipped MVP in four weeks. Cohorts are 30-40 people max, and instructors (mostly myself with some help from my friends) are people who have built and shipped real things, and we teach how to use AI as an assistive tool not your whole personality. Every project has a defined exit condition before anyone asks their tools to produce anything.

Every student presents what they built on demo day. Graduates join a permanent builders collective that grows with every cohort in a Discord group.

Ultimately we teach people: How to define what they are actually creating, what "done" looks like, how to maintain what they created and ultimately how to know when they have succeeded or failed.

Here's the syllabus:

Week 1: Problem decomposition and exit conditions. Creators have to write down exactly what they are going to present at the end of the course, how they think they will get there and what they plan to do with it after the course ends. Week 2: Prototype. First working version. Doesn't have to be pretty it just has to exist and meet the self defined exit condition. Week 3: MVP Testing and iteration. This is where the exit condition either holds or exposes what was wrong in week 1. Week 4: Demo day. You show however far you got to the cohort, and your plans for the future and get their feedback.

$99/week pay as you go, no long term lock in or special platform junk.

Students pick their project or I have 13 defined project types, each with a written exit condition on the site. Creators own everything they build outright, I don't have any platform dependency, no subscription, no equity taken. The only costs are the weekly fee and your time.

Cohort 0 starts May 4th. Full curriculum and exit conditions at givedirection.com.

Happy to talk through ideas and questions and happy to have anyone who wants to be an instructor let me know.

View the Vision, Mission and Roadmap here:

https://www.givedirection.com/vision.html

HN discount code for 50% off is COHORT-0

A quiet thread, for now.Start the conversation on HN ↗