# Jev demos show where AI decisions fit inside everyday software

> Eight Jev demos show how a typed AI decision can rank, select or filter within an existing app. We trace what the surrounding software still does and what remains untested.

By BIG CHANGE Editorial

Published: 2026-09-24T22:47:28.571Z
Updated: 2026-09-24T22:47:28.571Z
Canonical: https://bigchange.ai/blog/jev-demos-ai-decisions-everyday-software

![Charcoal illustration of a desktop monitor showing a generic web article with one passage highlighted in orange.](https://bigchange.ai/api/media/file/jev-find-in-page-hero-v1.png)
AI-generated conceptual illustration by BIG CHANGE.

Eight examples in [Matthew Berman’s September 24 video](https://www.youtube.com/watch?v=jGD_UR4wMJc) put TypeSafe AI’s Jev model into familiar tasks: cleaning a web page, finding a passage, ordering an inbox and selecting pieces of an interface. The demonstrations come from different builders. Their common move is to give a model a narrow judgment while the application gathers the input and carries out the result.

Each clip becomes more informative when the decision is separated from the work done in code or by another model. The possible error a user would notice matters as much as the model call. These are demonstrations and early tools; BIG CHANGE has not independently tested their accuracy or everyday reliability.

## The big change

- **What changed:** Builders are inserting a typed AI judgment into existing software interactions, from a browser shortcut to an inbox view. Jev returns a choice, score or probability; the application supplies the candidate material and acts on the answer.
- **Why it matters:** A useful decision can be made at the moment someone is reading, searching or sorting, without handing the whole task to a chat interface. The same design gives application owners responsibility for wrong rankings, hidden content and unintended actions.
- **What to watch:** The next evidence is task specific: whether these tools select the right passage, preserve essential page controls, rank important mail well and handle uncertain cases across ordinary use. A fast demonstration alone cannot settle those questions.

## A page cleaner shows the division of labor

[Kitze’s Unclutter browser extension](https://x.com/thekitze/status/2100595129874817340) is the clearest example. Its [project README](https://github.com/kitze/unclutter) says the extension extracts candidate page elements and asks Jev which are nonessential. Browser code then applies reversible hiding rules, saves them by page template and reapplies those rules without another model request. Users can pause the extension, keep an element visible or reanalyze a page. Cookie overlays may be hidden, but the extension does not press Accept or Reject for the user.

That is a consequential boundary. Jev can judge an element to be clutter; it does not own the browser, write consent choices or decide how long a rule survives. A practical evaluation would check whether navigation, accessibility controls, paywall notices or genuine consent options remain usable after cleanup. The project offers source builds and a bring-your-own-key setup, but its README does not provide an independent field study of those errors.

The [Made with Jev website detector](https://madewithjev.com/free-tools/ai-slop-detector), shown at [3:29 in Berman’s video](https://www.youtube.com/watch?v=jGD_UR4wMJc&t=209s), has a different pipeline. Its own method says a browser measures rendered styles, DeepSeek V4 Flash describes screenshots, Jev judges design and copy against a list of tells, and DeepSeek writes the short verdict. The site displays a 26% “slop” score for anthropic.com. That is the tool’s stylistic score, calculated from its own rubric. It does not establish how much of Anthropic’s site was written by AI, despite Berman’s inference in the video.

## Ranking information is a separate kind of decision

[Jonathan Unikowski’s inbox demonstration](https://x.com/jnnnthnn/status/2101399331115077760) proposes replacing reverse chronological order with a live importance ranking. His post says the feature is “coming to” Avec. It does not document a deployed inbox, the definition of importance or an independent measure of missed urgent messages. The application would still retrieve mail, present the order and decide whether anything is archived, labeled or sent; the demonstrated Jev role is prioritization.

[Shubham Saboo’s Needle extension](https://x.com/Saboo_Shubham_/status/2101576462042366114) makes the distinction easier to see. Saboo says Jev scores a page’s passages against the reader’s query and the extension highlights matching text in place. His [longer explanation](https://www.linkedin.com/posts/shubhamsaboo_introducing-a-new-way-to-find-f-withtypesafe-activity-7507513455138979841-CY8S) says Needle does not write an answer: the highlighted sentence already exists on the page. This changes what a find shortcut can search for, but a highlighted sentence can still be the wrong one. Testing needs queries whose correct passages are known, including pages with similar but conflicting statements.

[Burhan Usman’s clipping post](https://x.com/BurhanUsman/status/2101641842441732297) reports finding clips on a topic in a video longer than 90 minutes. In [the video segment at 8:34](https://www.youtube.com/watch?v=jGD_UR4wMJc&t=514s), Berman says the system is probably using a transcript; that is his inference, not a verified account of the pipeline. The post does not specify Jev’s exact input or output. A clipping application must still obtain the source material, establish time boundaries and produce downloadable clips. The post’s timing and cost are one builder’s report, without a published accuracy test or a breakdown of end-to-end work.

## What UI composition leaves to the application

[Chris Tate’s json-render experiment](https://x.com/ctatedev/status/2101022101750571357) assembles a user interface from application-owned options. The [project’s Jev documentation](https://github.com/vercel-labs/json-render/blob/main/apps/web/lib/jev/README.md) is unusually explicit: Jev chooses components and layout positions; code assembles and validates the specification; the renderer displays it. The playground’s business data is synthetic, and action handlers run on user interaction. The demonstration therefore shows a way to compose a bounded interface, not a model independently writing and deploying a website.

The two creative examples explore lower-stakes choices. [Matt DesLauriers’ color experiment](https://x.com/mattdesl/status/2100899669802963060) maps text prompts to palettes in a visual demo. [Stefan’s emoji experiment](https://x.com/heystefan_/status/2101369117496521042), shown at [9:52](https://www.youtube.com/watch?v=jGD_UR4wMJc&t=592s), ranks emoji against typed text. In each case an app displays selectable visual material. The posts show interaction ideas; they do not establish that the choices suit a brand, meet contrast requirements or work across languages and contexts.

TypeSafe’s [documentation](https://docs.typesafe.ai/introduction) explains why these examples have a family resemblance. Jev evaluates typed Choice, Score and yes/no questions against supplied state. Choice and Score return distributions and a confidence value that software can use in its own rules. The company [recommends testing thresholds on the actual task](https://docs.typesafe.ai/confidence); a confidence field is not an independent accuracy result.

The examples make a credible case for a design pattern: software can ask a small, timely question where a fixed rule is too brittle. Whether a particular tool earns a place in daily work depends on the errors it makes, what it exposes or hides, and whether the application gives people a way to recover. Those are properties of the whole workflow, beyond the model’s decision alone.

## Sources

- [Matthew Berman: 8 Jev Use Cases That Feel Like Cheating](https://www.youtube.com/watch?v=jGD_UR4wMJc) — The 10:38 video curates eight demonstrations from different builders. We reviewed its complete English auto captions and the chapter list. The 5:17–6:31 Zapier section is sponsored. The video is not independent product validation.
- [TypeSafe AI: Introducing System One Models & Jev](https://typesafe.ai/blog/introducing-system-one-models-and-jev) — Official early-access announcement and explanation of structured model decisions. Performance results are company tests with stated limitations; this article does not use them as a benchmark for the eight demonstrations.
- [TypeSafe AI: Introduction and Confidence documentation](https://docs.typesafe.ai/introduction) — Documents Jev's input state and Choice, Score and Noul question types. The companion confidence page explains that Choice and Score confidence is derived from their answer distributions and that action thresholds require testing on the user's task.
- [TypeSafe AI: Confidence](https://docs.typesafe.ai/confidence) — Explains the distinction between answer probabilities, derived confidence and task-specific operational thresholds. It does not establish accuracy for any demo in this article.
- [Made with Jev: AI Slop Detector](https://madewithjev.com/free-tools/ai-slop-detector) — The tool maker describes browser measurements, DeepSeek screenshot observations, Jev judgments and a separate DeepSeek verdict. Its 26% anthropic.com example is a score under the site's stylistic rubric, not evidence of AI authorship.
- [Kitze: Unclutter announcement and source](https://github.com/kitze/unclutter) — Original builder's README describes candidate classification by Jev, reversible local hiding rules, template reuse and user controls. The September 17 original post is https://x.com/thekitze/status/2100595129874817340. No independent live-field evaluation was found.
- [Jonathan Unikowski: Jev inbox prioritization](https://x.com/jnnnthnn/status/2101399331115077760) — The builder shows a live importance ranking and says the feature is coming to Avec. This establishes a demonstration and proposed release, not a deployed inbox or measured priority accuracy.
- [Shubham Saboo: Needle semantic find-in-page](https://x.com/Saboo_Shubham_/status/2101576462042366114) — The builder presents an open-source Chrome extension. His longer LinkedIn account says Jev scores existing page passages and the extension highlights them without generating a replacement answer; no independent retrieval evaluation is given.
- [Chris Tate: json-render and Jev experiment](https://github.com/vercel-labs/json-render/blob/main/apps/web/lib/jev/README.md) — Project documentation says Jev selects from app-owned UI candidates while code assembles and validates the specification and the renderer displays it. The original September 18 demo post is https://x.com/ctatedev/status/2101022101750571357. Playground data is synthetic.
- [Burhan Usman: topic-based video clipping demo](https://x.com/BurhanUsman/status/2101641842441732297) — The builder reports clipping a video over 90 minutes by topic. The post does not disclose complete pipeline, validated timing boundaries or clip accuracy; Berman's transcript-based explanation is an inference in his video.
- [Matt DesLauriers: Jev color experiment](https://x.com/mattdesl/status/2100899669802963060) — Original visual experiment pairs text with color palettes. It is evidence of a creative demo, not of suitability for production design or accessibility compliance.
- [Stefan: Jev emoji experiment](https://x.com/heystefan_/status/2101369117496521042) — Original post and Berman's 9:52 segment show emoji selection against typed text. No usability or cross-language evaluation accompanies the post.