---
title: "They were building in stealth for 2 years, I was building in stealth for 2 hours"
description: "Harsha Gundal made a Qwen model perform Jev’s visible trick in two hours, suggesting the useful interface may be much easier to copy than the proprietary model behind it."
author: "Paul Friedman"
kind: "link"
published_at: "2026-09-16"
url: "https://pau.fm/blog/2026/09/16/building-in-stealth-for-two-hours"
markdown_url: "https://pau.fm/blog/2026/09/16/building-in-stealth-for-two-hours.md"
canonical_url: "https://pau.fm/blog/2026/09/16/building-in-stealth-for-two-hours"
external_url: "https://x.com/harshagundal/status/2100310817522794812"
---

# [They were building in stealth for 2 years\, I was building in stealth for 2 hours](<https://x.com/harshagundal/status/2100310817522794812>)

> Happy to open source Qwen-2.5-1B-RLCD, 5x faster on-device inference for JSON workloads that need
> to be type-safe.

Seventeen hours after TypeSafe
[announced Jev](/blog/2026/09/15/introducing-system-one-models-and-jev), Harsha Gundal made a Qwen
model play Doom with a claimed 150-millisecond latency. This is not a reverse-engineering of Jev:
TypeSafe has not published its architecture or training method. It does reproduce the legible
product trick, reusing an ordinary model’s state to score closed choices in parallel instead of
typing JSON one token at a time.

That makes Jev more interesting and TypeSafe less defensible. A useful abstraction can be real while
a startup’s moat is not. If two years of stealth can be approximated with an inference mode for Qwen
before dinner, TypeSafe has to prove its training produces better decisions; speed and schema
validity have already escaped.