Post
CommunityJev Practices & Patterns
Don't Call It a Decision Model
TypeSafe AI blog post arguing Jev should be called a System One model rather than a decision model. It walks through a customer service workflow to show how to split work into small Jev judgments and code logic instead of one large Choice question.
- Format
- Post
- Reading time
- 2 min
- Published by
- Community
- Added
- Nov 2026
Highlights
- Published on the TypeSafe AI blog on Nov 3, 2026 and credited to Allie Laabs, DevRel.
- Argues that Jev makes judgments while code makes decisions, favoring the System One label.
- Traces the System One name to Daniel Kahneman's Thinking, Fast and Slow.
- Warns that one large Choice question stuffed with criteria prose leads to prompt gymnastics.
- Uses a customer service workflow with fraud, refund, retention, and answer branches as the example.
Questions about this entry
Short answers drawn from the linked project or page.
The post argues that Jev makes judgments while code makes decisions, so TypeSafe keeps the System One label instead of adopting the popular decision model name for the category Jev pioneered.
It was published on the TypeSafe AI blog on Nov 3, 2026, and is credited to Allie Laabs, DevRel at TypeSafe.
The post traces it to the concept Daniel Kahneman popularized in Thinking, Fast and Slow, where System 1 thinking is fast and intuitive and System 2 is slower and more deliberate.
It describes repeatedly tweaking long criteria prose inside a single large Choice question to cover each edge case, then failing again when an input nobody anticipated arrives.
A customer service workflow with branches for the fraud team, handoff to a person, instant refund, refunds team review, a retention offer, and a plain answer with no account action.
It recommends decomposing the decision into atomic judgments that genuinely need Jev and logic that code can answer exactly, then letting logic combine those answers into the decision.
It cites clarity of design, because you can see what actually flows through the system and which edge cases remain, and observability, because judgments and logic paths are kept distinct.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
Hacker News: Simon Willison's overview of Noul, Choice, and Score, and where a decision model fits.
benwerd53 pointsJev introduces a new shape of LLMRead the thread on Hacker NewsHacker News: swyx posts the Latent Space interview with Diogo Almeida.
swyx4 pointsJev: System One Models for Prod, Not God – With Diogo Almeida, CEO, TypeSafe AIRead the thread on Hacker NewsX: Diogo Almeida in five minutes
X: LLMs vs. Jev, clearly explained
More Jev resources like this
A Japanese experiment that builds a rules-based gomoku harness and replays a full Jev-versus-Jev game log to measure real per-move latency.
Jev Practices & PatternsArticle#community#article#games
Community
A skeptical Japanese replication that approximates Jev's speed by reading first-token logits from small LLMs, with Gemma3 and a Mario benchmark.
Jev Practices & PatternsArticle#community#article#open-models
Community
A use-case study comparing Jev, Mistral Small 4, and Gemini 3.5 Flash-Lite on rejecting unsuitable local-event listings, with individually tuned prompts, accuracy, response times, and costs.
Jev Practices & PatternsArticle#community#article#benchmarks
Community
A 2026 technical SEO audit price survey plus a measured case where a Jev pipeline crawled and judged 1,204 pages, producing 4,816 typed judgments in under three minutes for under $1 in model cost.
Jev Practices & PatternsArticle#community#article#pricing
Community
An independent baseline that reads typed option probabilities straight from a frozen Qwen3.5-4B's logits in one forward pass, reproducing Jev's interface pattern with open models rather than Jev's undisclosed model or training.
Jev Practices & Patterns#community#python#open-models
Community
TheoLeeCJ/SemIfPython
TheoLeeCJ/SemIfPythonApache-licensed, locally runnable Jev-style decision models (0.8B, 4B, 9B on Qwen3.5) with released weights, training code, a System One-compatible server, frozen eval suites, and a playground.
Jev Practices & Patterns#community#python#open-models
Community
jaredpalmer/kevPython
jaredpalmer/kevPythonFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.